UNIVERSITY FARM >/^JL TEXT-BOOK OF PALEONTOLOGY ^ö^m. MACMILLAN AND CO., Limited LONDON • BOMBAY • CALCUTTA MELBOURNE THE MACMILLAN COMPANY NEW YORK • BOSTON • CHICAGO DALLAS • SAN FRANCISCO THE MACMILLAN CO. OF CANADA, Limited TORONTO TEXT-BOOK OF PALEONTOLOGY EDITED BY CHAKLES R. EASTMAN, A.M., Ph.D. PROFESSOR OF PALEONTOLOGY IN THE UNIVERSITT OF PITTSBURGH AND CÜRATOR AT THE CARNEGIE MUSEUM, PITTSBURGH ADAPTED FROM THE GERM AN OF KARL A. VON ZITTEL LATE PROFESSOR OF GEOLOGY AND PALEONTOLOGY IN THE UNIVERSITY OF MUNICH SECOND EDITION REVISED AND ENLARGED BY THE EDITOR IN COLLABORATION WITH THE FOLLOWING-NAMED SPECIALISTS : R. S. BASSLER, W. T. CAL^LW, A. H. CLARK, H. L, CLARK, J. M, CLARKE, J. A. CUSHMAN, W. H. DALL, A. HANDLIRSCH, K. T. JACKSON, A. PETRÜNKEVITCH, P. E. RAYMOND, R. RUEDEMANN, C. SCHÜCHEBT, J. T. SMITH, F. SPRINGER, T. W. VAÜGHAN, C. D. WALCOTT \ I VOL. I. WITH ABOUT 1600 ILLUSTRATIONS MACMILLAN AND CO., LIMITED ST. MAETIN'S STREET, LONDON 1913 COPYRIGHT There i.s nothing good in tliis world but that wliicli is true." — Diderot. First Edition, 1900 Second Edition, 1913 1 I EDITOR'S PEEFACE TO THE SECOND EDITION A NEW Eiiglish editioii of von Zittel's Text-hook of Paleontology having been called for, advantage was taken of the opportunity to prepare a thorough- going revision of the first volume, in order that an adequate account might be incorporated of the new knowledge that has been gained during recent years. Towards this end, a number of specialists were invited to collaborate with the Editor in preparing a fresh treatment of the leading groiips of Inverte- brates, and the present work bears witness to the generous response that was made to this invitation. Many parts of the work have been entirely re- written, others have been emended, rearranged and enlarged, and the Classification in various places has been very considerably altered. The new work, therefore, cannot with either justice or propriety be called von Zittel's Text-book, being in effect a composite production ; and yet in scope nd style it is modelled after the well-known German original. The names of the diflerent collaborators appear on the title-page, and the ections that have been revised or rewritten are credited in the body of the work to the specialists responsible for them. To all of his collaborators the Editor desires to offer grateful acknowledgments, and to express the sense of his owii joersonal indebtedness to them for the large service they have endered, and for many individual courtesies. To his friend and former associate at Harvard, Doctor Eobert Tracy ackson, the Editor is under an Obligation greater than can be adequately icknowledged ; for besides having contributed practically a fresh account of he Echini, Dr. Jackson has carefully read the proofs of the entire work, and as offered in many places most valuable suggestions and emendations. Like everal of the other collaborators, also, he has furnished the Originals for a umber of new figures. The total number of fresh illustrations has thus been ensibly increased. It is hoped that the large amount of painstaking work hich has been bestowed upon the present treatise will be found to yield eturns in increased value and usefulness among students of Paleontology generally. CHARLES R. EASTMAN. I Carnegie Museum, Pittsburgh, Pennsylvania, June 5, 1913. Al'i'S LIST OF COLLABORATOKS FOR THE SECOND EDITION Dr. Joseph Augustus Cushman, Foraminifera. Dr. T. Wayland Yaughan, Adinozoa. Dr. EuDOLF EuEDEMANN, GmptoUtoidea. Mr. Frank Springer, Cystoidea, Blastoidea, Crinoidea. Dr. Arthur Hobart Clark, Post-Paleozoic Crinoidea. Dr. Hubert Lyman Clark, Asteroidea, Holothuroidea . Dr. Robert Tracy Jackson, Echinoidea. Dr. Ray S. Bassler, Bryozoa. Prof. Charles Schuchert, Brachiopoda. Dr. William Healy Dall, Pelecypoda, Gastropoda {jmrs). Prof. James Perrin Smith, Ammonoidea. Dr. Percy E. Raymond, TriloUta. Dr. Charles D. Walcott, TriloUta {pars), Arachnida (pars). Dr. Ray S. Bassler, Branchiopoda, Ostracoda. Dr. William T. Calman, Cirripedia, Malacostraca. Dr. John M. Clarke, Phyllocarida, Arachnida. Dr. Alexander Petrunkevitch, Embolobranchiafa, Myriapoda. Dr. Anton Handlirsch, Inseda. ORIGINAL AUTHOß'S PEEFACE Die englische Ausgabe meiner Grundzüge der Palaeontologie hat ein vom deutschen Original in verschiedener Hinsicht abweichende Gestalt erhalten. Der Herausgeber, mein Freund und ehemaliger Schüler Dr. Eastman, suchte mit meiner Zustimmung eine Anzahl der hervorragendsten Specialisten für die Bearbeitung einzelner Thierclassen zu gewinnen. Dadurch erfuhr das Werk eine gründliche und sachkundige Ueberarbeitung, welche sich namentlich im Detail vortheilhaft geltend macht und mancherlei Irrthümer der deutschen Ausgabe beseitigte. Für diese mühevolle und aufopfernde Arbeit bin ich den Mitarbeitern des Text-Book zu grossem Dank verpflichtet. Allerdings wurde durch die Betheiligung einer grösseren Anzahl von Autoren, deren Anschauungen in systematischen Fragen nicht immer unter einander und mit denen des Autors der deutschen Ausgabe in Einklang standen, die Einheitlichkeit des Werkes nicht unerheblich gestört und auch der ursprüng- liche Umfang verschiedener Abschnitte bedeutend überschritten ; allein diese Nachtheile dürften durch die sorgfältigere Durcharbeitung des eigentlichen Stoff'es reichlich ausgeglichen sein. Die Revision der Crinoideen hatte der verstorbene Herr Charles TFachsrmith, jene der Asteroideen und Echinoideen Herr fF. Percy Sladeii übernommen. Abgesehen von einigen Abänderungen, welche mehr terminologische als chliche Fragen betreffen, wurde in diesen Abtheilungen eine weit voll- tändigere Aufzählung und Charakterisierung der fossilen Gattungen durch- eführt, als in der deutschen Ausgabe. Weitergehende Umgestaltung erfuhr die Classe der Bryozoen durch Herrn E. 0. Ulrich. Die paläozoischen Formen eind von diesem ausgezeichneten Kenner mit einer Ausführlichkeit behandelt, welche nicht ganz mit der Darstellung anderer Abtheilungen in Einklang steht. Auch die Transferierung der bereits bei den Korallen abgehandelten Chaetetiden und Fistuliporiden zu den Bryozoen und die dadurch veranlasste doppelte Darstellung derselben ist eine Incongruenz, welche sich nur durch die Meinungsverschiedenheit über die zoologische Stellung dieser ausgestorbenen I Organismen entschuldigen lässt. Eine durchgreifende Umarbeitung haben die Brachiopoden durch Herrn Charles Schuchert erfahren. Während sich die deutsche Ausgabe mehr auf die viii TEXT-BOOK OF PALEONTOLOGY Werke und Anschauungen von Thomas Davidson stützt, folgt die englische Uebersetzung sowohl in der Auffassung der Gattungen und Familien, als auch in den systematischen Principien den neusten Arbeiten von James Hall, J. M. Clarke, und C. E. Beecher. Die systematischen Hauptgruppen sollen hier zugleich entwicklungsgeschichtlichen Phasen entsprechen und das ganze SjT'stem den Anforderungen des biogenetischen Grundgesetzes genügen. Von ähnlichen Gesichtspunkten wurden auch Professor Beecher bei der Bearbeitung der Trilobiten und Professor Hyaft bei jener der Gephalopoden geleitet. Es ist mir zweifelhaft, ob die Zeit zu einer durchgreifenden Reform der bio- logischen Systematik, bei welcher weniger morphologische und vergleichend- anatomische Merkmale, als embryologische und phylogenetische Gesichtspunkte im Vordergrund stehen, jetzt schon gekommen ist ; allein jedenfalls sucht die in Nord Amerika gegenwärtig herrschende Strömung auf einem neuen Weg zur Wahrheit zu gelangen und eine die genealogischen Beziehungen deutlicher wiederspiegelnde Systematik zu erzielen. Bei den Pelecypoden hat Herr Dr. JF. H. Dali die durch Neumayr eingeführten und in den Grundzügen mit einigen Modificationen angenommenen Gruppen durch seine eigene, auf langjährige Spezialuntersuchungen basirte Eintheilung ersetzt. Die Scaphopoden, Amphineuren, Gastropoden und Pteropoden wurden von Herrn Professor //. A. Pilsbry, die Crustaceen mit Ausnahme der Trilobiten und einiger anderer Gruppen von Professor /. S. Kingsley, und die übrigen Arthropoden von meinem langjährigen Mitarbeiter und Freund Professor S. H. Scudder in sachkundigster Weise durchgesehen. Für den wichtigen Abschnitt der Gephalopoden trägt Herr Professor Alpheus Hyatt die Verantwortlichkeit. Hier treten die Differenzen mit der deutschen Ausgabe am auffallendsten zu Tage, vertritt doch dieser Autor am entschiedensten die moderne Richtung in Amerika. Obwohl meine An- schauungen über verschiedene Grundprinzipien der Systematik, namentlich über Abgrenzung von Familien, Gattungen und Arten von denen meines amerikanischen Collegen abweichen, so glaubte ich doch einem so hervorra- genden Kenner der fossilen Gephalopoden bei der Bearbeitung des von ihm übernommen Abschnittes völlig freie Hand lassen zu müssen. Das Text- Book ist dadurch um eine Averthvolle Originalarbeit bereichert worden, welche viele bis jetzt noch nicht veröffentlichte Thatsachen enthält. Zu ganz besonderem Dank bin ich dem Herausgeber der englischen' Ausgabe Herrn Dr. C. R. Eastman verpflichtet. Er hat keine Mühe gescheut, das Werk mit den neusten Ergebnissen der paläontologischen Forschung in Einklang zu bringen und den Fortgang derselben zu fördern. Dr. KARL A. von ZITTEL. München, im Septe^yiber 1899. EDITOK'S PREFACE TO THE FIRST EDITION The Gnmd.ziige der Paldeontologie, which. forms the basis of the present work, was published in the spring of 1895, only a short time after the completion of the fifth and last vohime of Professor von Zittel's celebrated Handbuch der Falaeontologie. Of the latter, an excellent translation exists in French by Barrois ; but English-speaking students are without either an independent treatise on Paleontology or translation from any foreign work, which is comparable in scope and character to the writings of von Zittel. With the hope of supplying this deficiency the Editor undertook the task of rendering the Grundzüge into English. It was at first intended to bring out a strictly literal translation, but with the Author's consent this plan was modified in important respects which should be clearly understood by all. The chapters on Protozoa and Coelenterata stand here essentially as in the original, but nearly all the remaining chapters have been remodelled, enlarged, and brought as nearly as possible up to date by a selected body of experts. The greater part of the work is therefore a composite production, and from the nature of the case some incongruities in style and treatment are to be expected. For all the collaborators to have adhered to uniform limits of alteration and expansion would have been impossible. It will be found, therefore, that some portions of the revised text are not sensibly difFerent from the original while others are changed very radically, and a few chapters, notably the Molluscoidea, Mollusca, and Trilobites, are entirely rewritten. An eftort has been made throughout to adapt the text more especially to the needs of Anglo-American students, and the bibliographies have been enlarged with similar intent. For all changes in the Classification over the original tho revisers of the different sections are responsible ; but although radical departures have been made with the Author's sanction, one must by no means presume he is thereby committed to all the innovations which are set forth. How far and whether in all cases the system has been improved must be left for experience to determine. The Author's graciousness, however, in yielding his own prefer- ences on systematic points will be apparent on reading his annexed preface. Due acknowledgments are rendered the collaborators in the Author's X TEXT-BOOK OF PALEONTOLOGY preface, and also in footnotes at the end of the several chapters. Their names are enumerated below in the order of their respective sections, and the Editor begs to express at this time a sense of his profound appreciation of the Services that have been so generously rendered. For the many personal courtesies extended, he would return to each of them his sincere and hearty thanks. LIST OF COLLABORATORS Mr. Charles "Wachsmuth, Grinoidea, Blastoidea. Mr. W. Percy Sladen, Astero^wa, Echinowa. Dr. George Jennings Hinde, Vermes. Mr. Edward 0. Ulrich, Bryozoa, Ostracoda. Mr. Charles Schuchert, Brachiopoda. ])r. William H. Dall, Pelecypoda. Prof. Henry A. Pilsbry, Gastropoda. Prof. Alpheüs Hyatt, Gephalopoda. Prof. Charles E. Beecher, TrUohita. Prof. John M. Clarke, Eucrustacea (pars), Äcerata (pars). Prof. John S. Kingsley, Eucrustacea (pars), Äcerata (pars). Prof. Samuel H. Scudder, Insecta. The Editor is also greatly indebted to his friend Dr. John C. Merriam, who undertook the translation of the entire chapter on Mollusca, a very laborious work. Dr. Merriam's assistance has been further enlisted in the translation of the second volume, which will be devoted exclusively to the Yertebrates. Dr. August F. Foerste Avas kind enough to furnish a transla- tion of the chapter on Insects, and various friends have assisted in correcting proofs. For the compilation of the index the Editor is indebted to Miss Elizabeth B. Bryant, a former Student of his at KadclifFe College, and to his brother, Mr. David P. Eastman. CHARLES E. EASTMAN. Harvard University, SeiJtemher 15, 1899. CONTENTS INTRODUCTION PHYLUM III. Vermes (Worms) .... PAGE Definition and Scope of Paleon tology ..... 1 PHYLUM I. Protozoa . . . . . . . . .17 PHYLUM IL Coelenterata (Zoopliytes) . . . . • • .46 . 135 PHYLUM IV. ECHINODERMATA . . . . • • • .143 PHYLUM V. MOLLUSCOIDEA . . . • • • • .314 PHYLUM YL Mollusca ......... 421 PHYLUM VIL Arthropoda (Articiüates) ....... 690 INDEX 823 xi INTßODUCTION DEFINITION AND SCOPE OF PALEONTOLOGY Paleontology (Aoyos row TraXatCov ovTOiv) is the Science which treats of the life which has existed on the globe during former geological periods. It deals with all questions concerning the structure, Classification, relationships, descent, conditions of existence, and distribution in time and space of the ancient inhabitants of the earth, as well as with those theories of organic and cosmogonic evolution which result from such inquiries. Under the term of fossils are understood all remains or traces of plants and animals which have lived before the beginning of the present geological period, and have become preserved in the rocks. The criterion which determines the fossil character of organic remains is the geological age of the formation in which they occur, whereas their mode and State of preservation, or the fact of their belonging to extinct or to still living species, are merely incidental considerations. Although fossils have, as a rule, undergone more or less radical changes during the process of fossilisation, and are usually transformed into mineral substances, nevertheless, under exceptionally favourable conditions (as in frozen ground, amber, resin, peat, etc.), plants and animals may be preserved through geological periods in a practically unaltered state. Carcasses of mammoths and rhinoceroses entombed in the frozen mud-cliffs of Siberia, and inclusions of insects, spiders and plants in amber are none the less genuine fossils, in spite of their having sustained no trace whatever of mineral infiltration. A considerable number of plants and animals occurring as fossils in Tertiary and Pleistocene formations belong to still living species ; while, on the other hand, the remains of forms which have become extinct during historical times (Ehijtina, Alca, Didus, Pezophaps, etc.) can no more be classed as fossils in the true sense of the word than all such recent organisms as may chance to become buried in deposits now forming under the present prevailing orographic and climatal conditions. The changes which organic bodies undergo during the process of fossilisa- tion are partly chemical and partly mechanical in their nature.^ According as certain portions of the original substance are removed, or are replaced atom ^ White, Charles J., Conditions of preservation of invertebrate fossils. Bull. U.S. Geol. and Geog. Survey Territ., 1880, vol. v. , p. 133. Trahucco, G., La Petrificazione. Pavia, 1887. VOL I. B 2 ELEMENTS OF PALEONTOLOGY for atom by foreign matter, the result may be either carbonisation, decom- position, total dissipation, or petrifaction. Carbonisation is a deoxidising process taking place under water or with limited access of air, and especially common among plants. Fossil wood and other vegetable matter abound in peat, lignite and bituminous coal, the leaves being transformed into a thin flake of carbon, on which often the finest venation is still discernible. In some cases chitinous animal structures also become carbonised, as in insects, crustaceans and graptolites. Becomposition as a rule effectually destroys all organic carbon and nitrogen Compounds. With few exceptions, therefore, animals without hard parts, such as Worms, infusorians, naked mollusca, most hydrozoa, many anthozoa, and the embryos of vertebrates, leave no traces behind in the rocks. Hörn, hair, chitin and similar structures are likewise totally destroyed during the fossilisation process, while only under especially favourable conditions, as in ice or in frozen soil, muscular and epidermal tissues remain unchanged ; or eise, through the taking up of lime phosphate in argillaceous and calcareous deposits, undergo a sort of petrifaction, in which the finer structure is but little altered. ^ Even the conservable hard parts of animal bodies are deprived of their organic Compounds ; bones give up their fats and oils, and the shells of moUusks, echinoderms and crustaceans lose their pigments and soft sub- stratum. The hard portions, which first become more or less porous through loss of their organic constituents, next suffer the gradual disintegration of their inorganic Compounds, and finally undergo dissolution, reabsorption, or petrifaction. Petrifaction. — In this process foreign substances soluble in water (chiefly calcium carbonate and silica, more rarely pyrites, iron oxyhydrate and other salts) impregnate and completely fill all original cavities as well as those formed subsequently by decay. Chemical metamorphism takes place occasionally, when, owing to the decomposition of certain inorganic con- stituents, the original molecules become replaced by those of other substances. For instance, we find quartz pseudomorphs after calcareous tests and skeletal parts, and conversely, calcite pseudomorphs after silica, as in certain sponges. Wherever the space originally occupied by soft parts, as, for example, the interior of a shell or other hoUow body, becomes filled up with infiltrating ooze, while the shell itself or the enclosing wall decays, there is produced a cast of the interior, which in most cases (especially where the shell is thin, as in ammonites, brachiopods, certain moUusks and crustaceans) preserves an exact copy of the original form, and is susceptible of as accurate determination as the real object. Not infrequently fossil organisms leave molds or imprints of their shells or skeletons — very rarely of their whole bodies — in the rocks. Sometimes, indeed, their presence is indicated merely by tracks or footprinfs. Fossils are often distorted by mechanical agencies, such as faulting, folding, crushing, and other deformations of the country rock. Such cases require especial attention, and due caution must be observed in their determination. Paleontology and Biology. — Although the fossil remains of ancient life-forms yield but a fragmentary record of themselves, are almost never perfectly preserved, and are usually more or less altered in appearance, yet on the whole, they readily fit into place in the great framework of zoological ^ Reis, Otto, Über petrificirte Muskulatur. Arch. Mikros. Anat., vols. xli. xliv. lii. , — Dean, B., Preservation of muscle-fibres in sharks, Amer. Geol., 1902, vol. xxx. INTRODUCTION 3 and botanical classifications. Notwithstanding all their difFerences, their general structure is similar to that of recent organisms, and their Identification requires the most careful comparison with nearly related plants and animals. The methods of paleontological research do not differ from those employed by the zoologist and botanist, excepting, of course, that the paleontologist is restricted to those pärts which are alone capable of preservation, and must reconstruct the missing soft parts ideally from analogy with recent forms. It is, nevertheless, incumbent on the paleontologist to obtain all possible Informa- tion from the material such as it is, aided by every means he can devise ; and hence his investigations do not cease with an examination of the external, macroscopic characters, but must be extended to the finer microscopic and histological as well. In numerous instances paleontology has anticipated zoology and botany by important histological discoveries ; in the brauch of vertebrate comparative anatomy, for example, through the exhaustive study of conservable hard parts, such as the teeth, skeleton, dermal covering, etc., this science has been elevated to its present high Standard chiefly by paleon- tologists (Cuvier, Owen, Huxley, H. von Meyer, Rütimeyer, Marsh, Cope, Osborn and others). The principle of correlation of parts, first applied with such eminent success by Cuvier, according to which all parts of an organism stand in certain fixed relationships to one another, so that one organ cannot Vary without a corresponding Variation taking place in the others, is now worked out not only for the whole group of vertebrates, but for invertebrates as well ; and its elaboration is such that frequently a single bone, tooth, plate, carapace, shell-fragment and the like, is sufficient for us to form a tolerably accurate concept of the entire creature. It is therefore clear that in so far as paleontology has to deal with the study and Classification -of fossil organisms, it is no other than a part of zoology, comparative anatomy and botany, and hence may be very properly divided into Paleozoology and Faleo- lotany. Paleontology has vastly increased the subject - matter of the two biological sciences, has filled up innumerable gaps in the system, and has infinitely enriched our knowledge of the variety and complexity of plant and animal Organisation. In almost every class of both kingdoms where preserva- tion is possible, the number of fossil forms cohsiderably exceeds the recent. A natural Classification of the Foraminifera, sponges, corals, echinoderms, moUusks, vertebrates, and of the vascular cryptogams, cycads and conifers, would be utterly inconceivable without taking paleontological evidence into account, since in certain classes (brachiopods, cephalopods, reptiles, mammals) the number of extinct fossil forms may be ten, a hundred, or even a thousand- fold greater than the living, and this proportion is steadily increasing in favour of paleontology, as fresh discoveries are made in various parts of the World. Paleontology and Geology. — Although as a biological science paleon- tology does not differ essentially from botany and zoology, yet its connection with geology is none the less intimate, and consequently it has been cultivated quite as assiduously by geologists as by biologists. The material is brought to light almost wholly by geologists or by geological collectors, who obtain it from the stratified rocks of the earth's crust — that is to say, rocks which have been formed by the subaqueous deposition of sediment, or have been built up from detritus on dry land by aerial agency. The distribution of fossils throughout stratified rocks is by no means promiscuous, neither do all rocks 4 ELEMENTS OF PALEONTOLOGY contain the same species ; but on the contrary, each separate stratigraphical- complex, and frequently even single beds and layers, are characterised by certain particular assemblages of fossils. The older the rock, the-more strikingly different from recent organisms are its fossil remains ; the younger the forma- tion, the greater is their resemblance. Now, since experience shows that contemporaneous deposits which have been laid down under similar conditions (as, for example, in salt or in fresh water) contain identical or at least very similar fossils, the latter furnish us with an infallible guide, taken together with the local stratigraphic succession, for determining the relative age of a given formation. Furthermore, a knowledge of the fossils occurring in homo- taxial deposits enables us to reconstruct the various paleofaunas and paleo- floras which have existed on our planet at different periods in its history. Having determined the chronological succession of the clastic rocks by means of their superimposition and their characteristic or index-fossils, they may be divided up into still smaller series, each one of which is characterised by a particular assemblage of organic remains. In the main, then, paleontology is the ultimate foundation of historical geology. Excluding the oldest metamorphic rocks (gneiss, mica schists, phyllites, etc.) which are destitute of fossils, and concerning whose origin there is still great difFerence of opinion, the total thickness of the sedimentary rocks amounts to 20,000 — 30,000 metres. The building up of this prodigious pile of rock must have extended over an inconceivably long time, whose dura- tion cannot even approximately be estimated, since we are without data as to the rate of deposition in former periods, and since the beginning, culmina- tion and end of geological epochs cannot be correlated with astronomical events. Since, however, the earth has been inhabited in former times by very different creatures from those now living ; since successive paleofaunas and paleofloras follow one another everywhere in the same order ; and since, furthermore, in certain formations the greater part or even the total number of species appear and disappear in a body, so that one fauna or flora is replaced almost in its entirety by the next following ; it is obvious that the sedimentary rocks may be subdivided into a number of longer and shorter time measures, which may be designated by particular names. The beginning and end of such periods (group, System or formation, series or section, stage, Zone or bed) is usually indicated by local interruptions in the deposition, occasioned by variations in sea-level, volcanic eruptions, or by other causes ; and such disturbances are usually accompanied by changes in the flora and fauna. The now generally accepted subdivision of the secondary rocks is represented in the table on page 5, in which it should be noted that only the first three columns are of universal significance, while the last two apply only to European countries. The rocks of the Archaean Group amount to 40,000—60,000 metres in thickness. They belong to the oldest and longest period in the history of our planet, and are remarkable for their schistose and crystalline structure, as well as for the total absence of fossils. In order of stratigraphy, gneiss com- prises for the most part the oldest ; mica, chlorite, and talc-schists the middle ; and phyllites (primitive schists) the youngest division of this group. The so-called fossil organism, Eozoon, occurring in gneiss, has been proved to be of inorganic nature. ■ INTRODUCTION 5 Eras Periods 1 i Epochs (Etages) P4 1 Quarternary System Alluvium Recent Deposits Diluvium (Pleistocene) Postglacial Series Glacial Preglacial ,, Tertiary System 1 o Pliocene Upper or Siciliaii (öicilien) Middle or Astiaii (Astieii) Lower or Plaisaneiaii (Messinien) Miocene Pontian iSeries (Pontien) Sarniatian ,, (Sannatien) Tortonian ,, (Tortonien) Helvetian „ (Helvetien) Burdigalian Series (Burdigalien) Aquitanian ,, (Aquitanien) Oligocene Upper or Rupelian (Stanipien) Lower or Lattorfian (Tongrien) Eocene Bartonian Series (Bartonien) Parisian ,, (Lutetien) Ypresian „ (Ypresien) Landeniau ,, (Landenien) Mesozoic Group Cretaceous System Upper Cretaceous Danian Series (Danien) Senonian ,, (Senonien) Turonian ,, (Turonien) Cenomanian Series (Cenomanien) Lower Cretaceous Gault Series (Albien) Ai)tian and Urgonian Series (Aptien, Urgonien) Barremian Series (Barremien) Xeocomian and Wealden Series (Neocomien) Jurassic System Upper Jurassic (Malm) Tithonian and Portland Series (Portlandien) Kinimeridgian ,, (Kimmeridgien) Corallian ,, (Sequanien) Oxfordian „ (Oxfordien) Kellaway ,, (Callovien) Middle Jurassic (Dogger) Bath or Great Oolite Series (Bathonien) Bajeux or Inferior Oolite Series (Bajocien) Lower Jurassic (Lias) Toarciaii Sfiies (Toarcien) Charmouthiau Series (Charmouthien) Sinemurian ,, (Sinemurien) Hettangian ,, (Hettangien) Rhaetic Stage (Rhaetien) Noric „ (Norien) Carinthian ,, (Carnien) Triassic System Upper Triassic (Keuper) Middle Triassic (Muschelkalk) Ladinian Series (Ladinlen) Virglorian ,, (Virglorien) Lower Triassic (Bunter Sandstein) Werfenian series <'äyEj°' ^^Bu Permian System (Dyas) Zechstein Rothliegendes Thuringian Series (Thuringien) Saxonian or Penjabian (Saxonien) Artinsk or Autunian (Autunien) Carboniferous System Coal Measures Uralian or Steplianian (Ouralien) Muscovian or Westphalian (Moscovien) Lower Carboniferous (Culm) Dinantian Series or Culni (Dinantien) Devonian System Upper Devonian Famennian Series (Famennien) FrasniaTi ,, (Frasnien) Middle Devonian Givetian ,, (Givetien) Bifelian Series (Eifelien) Lower Devonian Coblenzian Series (Coblentzien) Gedinnian ,, (Gedimiien) Silurian System Silurian Ludlow Series ^ Wen lock „ WGothlandien) Llandovery Series j Ordovician (Lower Silurian) Bala or Caradoc Series ) Llandeilo „ WOrdovicien) Arenig „ ) Cambrian System Upper Cambrian is'Är'f^ (rot»"-™) Middle Cambrian Menevian Series (Acadien) Lower Cambrian Harlech Series (Georgien) Arch- aean Group Primitive Schist System Gneiss System Phyllite (Primitive Schists, Mica and Chlorite Schists, etc.) 1 Gneiss Precambrian Series (Algonkien) 6 ELEMENTS OF PALEONTOLOGY The Faleozoic or Primarij Group comprises the Cambrian, Ordovician, Silurian, Devonian, Carboniferous and Permian Systems, each of which is made up of a great number of series, stages and zones. In the Cambrian crustaceans, trilobites, brachiopods and worms predominate, associated with a few echinoderms, coelenterates, sponges and poorly preserved algae. In the Silurian system most classes of the animal kingdom are represented with the exception of amphibians, reptiles, birds and mammals, while the flora still consists of algae. Marine invertebrates are very abundant, especially crus- taceans, moUusks, echinoderms and coelenterates, while only a few frag- mentary fish-remains indicate the presence of vertebrates. All the species and nearly all the genera have since become extinct, and belong for the most part to extinct families and Orders. During the Devonian, Carboniferous and Permian Systems, the same classes of animals continue as a body, but are represented by frequently different families and genera. Fishes develop a great variety of forms in the Devonian, amphibians (Stegocephalia) make their appearance in the Carboniferous, and reptiles in the Permian. The flora consists chiefly of vascular cryptogams, together with a few conifers and cycads. The Mesozoic Group comprises three Systems — the Triassic, Jurassic and Cretaceous. Many of the widely distributed Paleozoic types (Tetracoralla, graptolites, crinoids, cystids, blastoids, brachiopods, trilobites) have either wholly or in greater part disappeared, while others (cephalopods, lamellibranchs, sea- urchins) are replaced by very different genera and families. Vertebrates are remarkable for the gigantic size attained by amphibians (Labyrinthodonta) and many reptiles, as well as for the wonderful variety of the latter. Birds appear for the first time in the Upper Jurassic (Archaeopteryx), and mammals towards the close of the Triassic, being represented by diminutive, probably marsupial types. During the Triassic and Jurassic periods, vascular crypto- gams, conifers and cycads remain the dominant plant-forms, dicotyledons not occurring until the middle Cretaceous. The Cenozoic Group comprises the Tertiary and post-Tertiary or Quaternary Systems. Among the invertebrates, ammonites, belemnites, Rudistae and most of the Crinoidea have now passed away. Amphibians and reptiles have greatly declined, and, like the invertebrates, are represented by still living Orders. On the other band, birds, and particularly mammals, attain a wide distribution ; the latter class branches out in such manifold variety, and experiences such rapid development during Cenozoic time, that it alone furnishes us with the principal index-fossils of this era. From now on the flora consists chiefly of dicotyledonous plants. Paleontology and Physical Geography. — Not only do fossils con- stitute the very foundation of historical geology, but they furnish us in addition with valuable inform'ation respecting the origin of the rocks in which they occur, the former distribution of land and water, climatal conditions, and the facts of geographical distribution in former periods. By means of analogy with recent species we are able in most cases readily to determine whether fossil forms pertain to land, fresh, brackish or salt water species, whence it is apparent under what conditions the strata were deposited. The distribution of marine and fresh-water formations helps us to certain conclusions respecting the extent of former seas and land areas. Deep-sea, shallow water, and littoral deposits are readily distinguishable by means of INTRODUCTION 7 their fossil organisms. By fossils, also, even the climatal conditions of former periods are indicated with great fidelity. The luxurious and uniform development of cryptogams over the face of the globe during Carboniferous time presupposes a warm, moist climate, little varying with latitude ; tropical dicotyledons occurring in the Cretaceous and Tertiary deposits of Greenland,or coral-reefs extending into high latitudes during the Paleozoic era, prove with equal certainty the prevalence of a milder climate and higher oceanic temperature in earlier times ; while again, the remains of reindeer, the lemming, musk-ox, polar fox, and other arctic animals in the diluvium of Central Europe testify to a period of glaciation with reduced mean annual temperature. The geographical distribution of fossil organisms proves that the regions and provinces occupied by recent plants and animals are to a certain extent identical with those existing in the Tertiary, and that life has been subject to the same distributional laws in the past as in the present. Nearly all recent forms are the obvious descendants of extinct creatures which formerly occupied the same region. For example, the fossil mammals, birds and reptiles of Diluvial time in Europe, Asia, Australia, North and South America, are scarcely distinguishable from forms now inhabiting the same continents. The ancestral homes of marsupials and edentates were perpetuated in Australia and South America until as recently as the Diluvial epoch, and during the later Tertiary, Europe, Asia and America formed but a single zoological province, inhabited by the ancestors of forms now living in the northern hemisphere. An understanding of the physical conditions which have governed the perpetuation of recent plants and animals in their respective provinces (succession of similar types) would be utterly impossible without a knowledge of their distribution in former times. In like manner, our know- ledge of the distribution of land and water, of prevailing climatal conditions, oceanic currents, etc., of earlier periods depends chiefly upon evidence derived from fossils. Paleontology and Embryology. — To trace the development of living plants and animals through all stages from the one-celled egg onward to final dissolution, is the task of Embryology and Ontogeny. At the present moment, botanists and zoologists are devoting their most scrutinising attention to embryological investigations, which latter accordingly exert a powerful influence on the progress of biology, and particularly on the Classification. The fact that every individual, species and genus of a whole group of plants and animals passes through nearly the same course of development, at least in the primary stages, and that all embryos belonging to a given order or class resemble one another so closely, up to a certain stage, that they cannot be told apart, has revealed unexpected affinities among forms differing very considerably in the adult stage. Cirripedes, for example, which were formerly mistaken for shell-bearing mollusks, develop from the same Nauplius-larvae as the Copepoda, Branchiopoda and Ostracoda, although the mature individuals belonging to these Orders of crustaceans possess but little common resemblance. Likewise, the whole group of vertebrate embryos can hardly be distinguished from one another in the earliest stages, and only very gradually assume the characteristic features pertaining to class and order. The results of embryological inquiry have a most important bearing on paleontology. Numerous fossil forms are known, which, in comparison with recent related organisms, exhibit embryonic, or at least larval or adolescent 8 ELEMENTS OF PALEONTOLOGY characteristics. Examples of such primitive or emhryonic fypes are especi- ally common in vertebrates, for the reason that here the skeleton becomes ossified very early in life, and hence the immature stages of the recent can be directly compared with adult fossil forms. Now, Observation has shown that in most of the older fossil fishes and reptiles, the vertebral column never passed beyond an embryonic stage, but remained in a cartilaginous or incompletely ossified condition through life. The Paleozoic amphibians (Stegocephalia) probably breathed by means of both gills and lungs through- out life, whereas most recent amphibians lose their gills comparatively early (Caducibranchia), and breathe wholly by lungs. Many fossil reptiles and mammals retain certain skeletal peculiarities permanently, while allied recent forms exhibit them only in embryonic stages. The skull in most of the older fossil reptiles and mammals closely corresponds in form and structure with that in embryos of recent related types. In the oldest fossil artiodactyls the palm-bones are all completely separated, while in recent ruminants this division continues only during the embryonic stage, being followed by a fusion of the two median metapodals, together with a reduction of the laterals. Among invertebrates, also, fossil embryonic types are by no means uncommon. The Paleozoic Belinuridae find their counterpart in the larvae of the common Limulus ; many fossil sea-urchins are characterised by linear ambulacra, while recent related forms, although developing petaloid radii in the adult stage, pass through the linear phase during adolescence. Many fossil crinoids re- semble the young of the living genus Antedon ; and, according to Jackson, recent echinoids, oysters and pectens exhibit in their nepionic stages certain characters peculiar to the adults of Paleozoic genera. The so-called fossil generalised or comprehensive types, which unite in one and the same form characters which, in geologically later, or recent descendants, have become distributed among different genera and families, are in reality merely primitive or immature types which have stopped short of the higher differentiation attained by their descendants. Generalised types always precede more highly specialised ; and properties that were originally distri- butive among older forms are never reunited in geologically younger species or genera. Trilobites, amphibians and reptiles of the Paleozoic and Mesozoic eras, and early Tertiary mammals belong almost exclusively to the category of generalised types. In certain groups of vertebrates, and especially of mammals (Ungulata, Carnivora), the chronological succession of genera is so closely paralleled by the successive stages of development in the life-history of their descendants, that to a certain extent the ontogeny of the individual is a representment of a long chronological series of fossil forms. This truth furnishes a strong foundation for the hiogenetic law, enunciated in various terms by GeofFroy St. Hilaire, Serres, Meckel, Fritz Müller and others, and recently more precisely formulated by Haeckel, as follows : The developmental history or ontogeny of an individual is merely a short and simplified repetition or recapitulation of the slow (perhaps extending over thousands of years) process of evolution of the species and of the whole brauch. The biogenetic law has since been found to hold true not only for verte- brates, but also for invertebrates, including even wholly extinct types. In ammonites, for instance, the primary or innermost whorls always difFer frora the outer in their greater simplicity of suture, and in their lesser ornamenta- INTEODUCTION 9 tioii. Very ofteii a correspondence is observable with geologically older forms ; and it is a well-known fact that all ammonites pass through early stages whicli resemble, at least so far as chambering of the shell is concerned, Paleozoic goniatites. A comparison of the inner whorls of an ammonite with its corre- sponding goniatitic form, or with older ammonites, seldom fails to reveal ties of kinship not otherwise discernible. Beecher has shown that nearly every stage in the development of arm-supports in recent brachiopods corre- sponds to conditions of the adult in some fossil genus ; and further, that the chronological succession of the latter is to a certain degree identical with the successive ontogenetic stages of recent forms. The relation of rudimentary or degradational organs occurring in recent forms to those of the fossil ancestors of the latter is of extreme significance. By rudimentary organs are meant certain structures (as, for example, limbs, parts of limbs, organs of sense, respiration, digestion, reproduction, etc.), which are still indicated by atrophied remains, but whose physiological functions, and hence their Utility to the organism, have wholly disappeared. Rudimentary organs are, as a rule, either normally developed in an embryonic stage, or at least more strongly than in the adult individual, owing to a process of degeneration, or retrogressive development. The fossil progenitors of forms possessing vestigial structures are almost always characterised by a füll development of the respective parts. The lateral metacarpals and metatarsals in the horse and most ruminants, for example, are indicated only by rudimentary side-splints ; but in an embryonic stage they are much more strongly developed, and in related fossil forms they occur as normal bones, carrying toes like the other metapodals, and serving for locomotion and Support. The wrist and metacarpal bones in birds have also sufFered degenera- tion, as is evident from a comparison with embryos and wdth older forms (Archaeopteryx), which exhibit a much higher development. In like manner, the teeth of birds have also become degenerated. In only a few forms (parrots, ostriches) are faint dental ridges discernible during embryonic stages ; but in all known Mesozoic birds the teeth are well developed and remain functional throughout life. Similarly, teeth are developed during embryonic stages in the baleen whale, but subsequently become atrophied ; while in the older fossil Cetacea teeth are always present. Other instances of this nature are to be met with in great profusion, both among vertebrates and invertebrates. The biogenetic law is, however, not infrequently obscured, for the reason that two closely related forms may not develop in exactly the same manner ; embryos of the one type may be afFected by peculiar accelerating impulses which are not shared by those of the other, and in consequence the first may pass through certain stages very rapidly, or may even omit them altogether. In this way the historical or palingenetic record contained in the development of every individual may be to a large extent veiled, suppressed or rendered unintelligible ; and this phenomenon of inexact parallelism (coenogenesis) is especially common in highly differentiated types, where the embryo passes through a multitude of phases. Paleontology and Phylogeny. — While conceding that by means of embryological investigations zoologists and botanists are able to trace the gradual development and differentiation of an organism through all its various stages, and thereupon to construct a tree of descent (phylogeny) founded upon 10 ELEMENTS OF PALEONTOLOGY the successive phases of grovvth, nevertheless such hypothetical genealogies can only be relied upon as truthful when they are substantiated by paleonto- logical facts. And only in cases where the difFerent ontogenetic stages are represented by corresponding fossil primitive or generalised types, which appear in the same chronological order, and clothe the supposititious ancestral tree with real forms, can the truthfulness of the latter be said to have been established. This requirement paleontology is from the nature of things unable to satisfy except in a few instances ; but a multitude of other facts testifies to the blood-kinship between morphologically similar fossil and recent organisms, and points to the direct descent of the younger from the older forms. Geology proves conclusively that of the numerous floras and faunas which lie buried in the rocks, those which are most nearly of the same geological age bear the greatest resemblance to each other. It often happens that species and genera occurring in a given formation reappear in the next following with scarcely any perceptible changes, so that the doctrine of the gradual trans- formation and transmutation of older forms is irresistibly forced upon one, while the faunas and floras of later periods assert themselves as the obvious descendants of the more ancient. Other weighty evidence for the progressive evolution of organisms is afforded by fossil transitional series, of which a considerable number are known, notwithstanding the imperfection of the paleontological record. By transitional series are meant a greater or less number of similar forms occurring through several successive horizons, and constituting a practically unbroken morphic chain. Often the differences between individuals belonging to diff'erent periods are so slight that we can hardly assign to them the value of a variety. But let a number of such mutations occur in succession, the end-members of the series become finally so divergent as to constitute distinct species and genera. The most striking and most numerous examples of transitional series naturally occur in types peculiarly well fitted for preservation, such as mollusks, brachiopods, sea- urchins, corals and vertebrates. Particularly remarkable among mollusks are the closely linked transitional series in ammonites. Among vertebrates transmutation proceeded far more rapidly than among invertebrates, and accordingly, the successive members of a series are usually so divergent as to require their assignment to separate genera. With increasing abundance of paleontological material, the more numerous and more complete are the series of intermediate forms which are brought to light. But the more extended our knowledge of transitional series the greater is the difficulty we encounter in defining our conception of species. While the older disciples of the Linnean and Cuvierian schools contended that each separate species was created with a certain definite sum of fixed characters, and remained incapable of any extensive modifications ; on the other band, those holding to the Darwinian theory of evolution look upon varieties, species, subgenera, genera, families, Orders, classes and phyla merely as arbitrary yet useful and convenient distinctions, corresponding to the state of our information at the present time ; it being assumed that by means of gradual transmutation during the course of ages all organisms have become evolved from a single primitive cell, or from a few primitive types. According to the Linne-Cuvier doctrine, a species is composed of individuals which are directly descended from one another, or from common ancestors, INTRODUCTION 13 and which resemble their progenitors as much as they resemble each oti^es Membeis of one and the same species interbreed, but individuals belongin^ to different species do not cross, or when they do, produce infertile or iraperfectly fertile offspring. According to the theory of descent no sharp specific distinctions can be drawn, but all individuals are assigned to the same species which share a number of essential features in common, and which are not connected with neighbouring groups by means of intermediate types. It is piain that this definition is open to considerable laxity of interpretation, and inasmuch as the direct descent of individuals belonging to a given species cannot always (in paleontology never) be determined on experimental grounds, systematists are rarely agreed in regard to the precise limitations of species, genera and families. The doctrine of the invariability of species received powerful support from the cataclysmic theory of Cuvier, which maintained that each period in the earth's history is marked by distinctively characteristic faunas and floras ; that no species is common to two successive periods ; that tremendous convulsions of nature (catadysms) occurred at the close of each cycle, and annihilated the whole organic world ; and that by means of special creative acts, the renovated earth became time and again populated with new animals and plants which bore absolutely no connection with either previous or subsequently introduced types. Cuvier's cataclysmic theory raay be regarded at the present day as com- pletely overthrown, inasmuch as the modern school of geology, following the leadership of Sir Charles Lyell, has demonstrated conclusively that the earth has proceeded from one stage to another during the course of its development only with the utmost slowness ; that the same forces and laws which regulato the World of to-day have operated likewise in primeval times ; and that geological periods are by no means abruptly set off from one another, but are linked together by innumerable transitional stages. The theory of the descendant origin of organic forms, which was advanced as early as 1802 by J. B. Lamarck and Geoffroy St. Hilaire, and was supported by Goethe, Oken and Meckel in Germany, kept winning continually more adherents, yet it was not until the latter half of the nineteenth Century that its universal significance was insisted on by Charles Darwin and his school. Paleontology, as already remarked, contributes a great deal of extremely weighty evidence in favour of the theory of descent ; the series of intermediate forms, often traceable through several successive formations ; the presence of primitive and generalised types ; the parallelism between ontogeny and the chronological succession of related fossil forms ; the similarity between floras and faunas of approximately the same age ; the correspondence in the geographica! distribution of recent organisms with that of their progenitors ; and a host of other facts are explicable only by means of the theory of descent. The causes of Variation and transformation were attributed by Lamarck chiefly to the use and disuse of organs ; secondly, to the effect of changes in external conditions ; and lastly, to a supposed inherent tendency toward Variation and perfection existing in each individual. According to Lamarck, new characters brought about by these influences are transmitted to descend- ants through inheritance, and become permanently established in the race. Geoffroy St. Hilaire maintained the same principles on the whole, but ascribed the chief causes of Variation of species to the infiuence of environment. 12 ELEMENTS OF PALEONTOLOGY The Darwinian theory of natural selectioii is based upon the property common to all organisms of acquiring ancestral characteristics through heredity, and of transmitting them in turn to their progeny ; and also on the adapta- bility of organisms to particular external conditions, by means of which variations are brought about. Since in the struggle for existence only those individuals which are the best adapted — that is to say, those possessing the most advantageous modifications — survive, nature is continually exercising, according to Darwin, a most rigorous selection which operates toward the increase and perfection of useful variations. Through the constant accumula- tion of originally slight yet serviceable modifications, and through the perpetual transmission of the same from one generation to another, there are produced first of all new varieties,then species, and eventually genera, families and Orders. The zoological and botanical classifications are, according to Darwin, merely an expression of genealogical facts, exhibiting the remoter and closer ties of consanguinity which exist among difFerent organic forms. Darwin's explanation of the origin of species by the addition of the agency of natural selection to the Lamarckian factors of Variation and inheritance found in Wallace, Huxley, Haeckel and others, zealous and ingenious supporters, although on other sides it encountered vehement Opposition. Moritz Wagner regarded free intercrossing as an insurmountable obstacle to the establishment of new modifications, and contended that the isolation of a few individuals, a condition which w^ould occur most frequently during migrations, was a necessary postulate in accounting for the origin of each new variety or species. As will be stated presently, the principle of isolation, slightly modified, has been applied by other writers. Bronn, Nägeli and A. Braun raised the objection to Darwin's theory of natural selection that many organs are entirely useless to the individual, and therefore natural selection, which depends upon the principle of utility, could neither have produced such organs nor could have modified them in any way. Nägeli assumed that, in addition to natural selection, a certain resident tendency toward perfection, inherent in every individual, takes part in conditioning the growth of morphological characters. Every Variation brought about by external or internal agencies is at once in the nature of a differentiation, a Step forward in the division of labour, and consequently an advancement. Weismann endeavoured in a similar manner to Supplement Darwin's theory of selection by his hypothesis of the continuity of germ-plasm. According to Weismann, germ-matter is of itself capable of producing all variations that are useful to an organism. Only that which exists in the original plasm or in the sexual elements as embryonic rudiments can be transmitted to off'spring and become further acted upon and developed by natural selection, according to Weismann's theory. The continuity, that is to say, the perpetual trans- mission of a portion of the germ-plasm from parent to ofFspring, forms a necessary postulate to the theory of descent. Weismann originally attributed only a subordinate influence to the action of physical environment as a cause of variations, and particularly denied the inheritance of acquired characters. But in his later writings, he is inclined to admit that somatic variations due to environmental influences may be transmitted to the ofFspring, and endeavours to explain this with the help of his germ-plasm hypothesis. Thus he approaches in a way the opinion of his opponents, the so-called Neo-Lamarckian school (represented by Herbert INTEODUCTION 13 Spencer, Cope, Hyatt, Osborn, Semper, Claus, Roux and others), which ranges itself more and more on the side of Lamarckian ideas, and ascribes to the use and disuse of Organs, and to external conditions, a very considerable influence in effecting the transformation of organic forms. While, on the one band, Semper, Locard and Clessin undertake to prove the direct action of environ- ment on molhisks in a number of instances ; on the other band, Cope, Osborn, Roux and others, emphasise the effect of use and disuse, and abundance or scantiness of food-supply. Adequate nourishment and exercise increase the development of a given organ, while physical conditions determine its form. Since like causes produce like efFects in the animate as well as in the inanimate World, it is obvious that similar organs must be developed in a variety of plant and animal forms wherever they are subjected to similar external conditions, and especially to the same physical agencies. A convenient explanation is thus found for the phenomena of parallelism, or " convergence," which are in nowise related to one another by inheritance. The analogous swimming- organs of fishes, ichthyosaurians and whales, or the analogous limb-structure in long-legged ruminants, the horse, elephant and Carnivora, are due to adaptation to external conditions and to use ; the same explanation also accounts for the like form of sternum in bats, birds and Pteivsauria, or for the spindle-shaped body characteristic of most rapid-swimming fishes, reptiles and aquatic mammals, or for the similar form of jaw possessed by marsupials and varlous Orders of Placentalia. These are all instances of parallelism, in which it often happens that two fundamentally different forms acquire the same outward shape, or become provided with similar or analogous organs. Kineto- genesis, or the process of a gradual transformation of parts, especially parts belonging to the internal skeleton, skull and limbs, is very ingeniously interpreted by Cope as having been accomplished in mammals through the agency of mechanical conditions, use and food. The same author has also traced the line of progressive modification in fossil genera as exemplified by numerous series of intermediate forms. In sharp contrast to all these opinions is the " mutation theory " of de Vries. The latter attempts to show that new species of plants are formed by what he calls mutations. It should be noted that this term is used in a different sense from the same word as mentioned on p. 10, being equivalent to saUation, used previously for the same thing. Mutations, in de Vries's sense, are more or less strongly marked deviations from the normal type, appearing rather suddenly ; and de Vries claims that only these are capable of being bred true by pedigree-culture, and that they alone lead to the origin of new species. What he actually did was to demonstrate that it is possible, by pedigree-culture, to produce true breeding forms (species) out of mutations, but he failed to see that the essential factor in this process is not the quality of the material he worked with {i.e. the mutations), but that it is the pedigree- culture, and that this corresponds to the well-known factors of selection and Isolation. The latter principle, originally introduced, as has already been stated, by Moritz Wagner, has recently been put forward by other writers (Baur, Ortmann, Gulick, etc.) as a factor which causes the differentiation of one species into several co-existing species (" process of speciation," O. F. Cook). While it is admitted that the Lamarckian factors of Variation and inheritance and the Darwinian factor of natural selection, are real and actual, it is 14 ELEMENTS OF PALEONTOLOGY evident that these are not sufficient for the understanding of the whole of the evolutionary process. They are capable of explaining the transformation of one existing form into one other form, but fail to account for the fact that often tvvo or more different forms have originated from a single ancestral type. Isolation, biological or ecological Separation, or habitudinal segregation, are synonymous terms applied to a fourth factor, which is important for render- ing the process of speciation more intelligible. These terms signify that the descendants of one ancestral form living amid a definite set of ecological (or environmental) conditions, begin to adapt themselves to different sets of conditions, as a result of which they become ecologically separated or segregated. Each group of descendants consequently becomes subject to different infiuences of environment, and in responding to such develops along different lines, the diverg'ence becoming finally so great as to be of specific value. According to the view that has just been stated, it is necessary to recognise that the whole process of evolution is a very complex one ; that it is the ultimate outcome of a number of factors, each of which has its own special efficacy, and may be sometimes antagonistic to the others ; and that of the various factors engaged the following four are the most potent and most essential : Variation, inheritance, natural selection and Separation. This view is perhaps to be regarded as the most satisfactory explanation of the organic World and its upbuilding that has yet been put forward. Nevertheless, though it cannot be gainsaid that the four operative principles just mentioned are actively at work, it is difhcult sometimes to trace' their causes. This is particularly true of those factors known as Variation and inheritance. As to the former of these factors, the rival hypotheses of the Lamarck-Darwinian and of the Weismannian school are contradictory with reference to the cause of inheritable Variation. With regard to the cause of inheritance, important discoveries, such as the Mendelian law, have been made, but these are too far- reaching to permit of a satisfactory account in limited space. Life-Period and Extinction of Species. — Observation shows that different organisms are by no means equally susceptible to Impulses received from the outer world. Many fossil genera remain almost wholly unchanged throughout a number of formations (Foraminifera, Cidaris, Nautilus, Lingula, Terebratula, Insectivora), and hence may be designated as persistent or conservative types, in contradistinction to variable types. The latter pass through rapid changes at the beginning of their career, develop a great variety of forms, and send out branches and off-shoots in all directions up to a certain point ; they may then die out af ter a comparatively short period of ascendency (Nummulites, Graptolites, Cystids, Blastoids, Tetracoralla, Perischoechinoida, Trilobitae, Rudistae, Ichthy o- sauria, Pterosauria, Dinosauria, Amblypoda, Toxodontia, etc.), or in some cases may even continue on to the present day with undiminished vitality (Spatangidae, Clypeastridae, many land and fresh-water mollusks, crabs, lizards, snakes, ruminants, apes). Not infrequently types that were originally variable pass over gradually into persistent ; their power of adaptation dwindles, they grow less plastic, become incapable of sending off new varieties, species or genera, and as the less vigorous of their number become worsted one after another, they finally stand out as isolated relics of antiquity {Isocrinus, Hatteria, 2'apirus, Equus, etc.) in the midst of rehabilitated surroundings. A one-sided develop- ment in a certain direction, excessive size, abnormal (hypertrophic) peculiarities, or too high specialisation of organs, is as a rule injurious to the form and INTRODUCTION 15 leads usually to its extermination. Many groups remarkable for their extreme differentiation (Dinosauria, Pterosauria, Amblypoda, Toxodontia, etc.) have become extinct probably for this reason, since, having advanced so far in a single limited direction, adaptation in other directions was no longer possible. Persistent types seldom produce a large number of species during a single geological period ; types that start up suddenly and proceed to vary rapidly as a rule soon die out ; while groups that develop slowly and steadily usually contain in their growth the promise of great longevity. Some very ancient types have persisted to the present day in highly saline lakes, or in salt paus, in acid, alkaline, very cold or otherwise unnatural situations. These represent dominant types of past ages which, vigorous and adaptable, when forced by internal specific pressure due to the enormous increase in the numbers of individuals were able to invade and adapt them- selves to physically and chemically unfavourable localities. The subsequent development of other types, younger and more vigorous, has extirpated them from all the more desirable situations, though these types have not proved sufficiently adaptable or vigorous entirely to exterminate them. The fauna of the ocean deeps and of biologically unfavourable situations generally is, therefore, a curious composite of the more vigorous and adapt- able types of animal life from the Cambrian to the present day, including f orms which were dominant in all earlier epochs, as well as forms derived directly from recent ancestors. For the extinction of many plants (Sigillaria, Lepidodendron, Cordaites) and animals (Blastoids, Tetracoralla, Trilobites, Ammonites, Rudistae, Ichthyosaurs, etc.) of former periods no adequate explanation has as yet been found. Changes in external conditions, especially such as regards the distribution of land and water, climatal conditions, saltness of the water, volcanic eruptions, paucity of food-supply, the encroachments of natural enemies, and diseases, may have led to the extinction of certain forms, but such conjectures signally fail to account for the disappearance of an entire species or particular groups of organisms. Oftentimes extinction seems to have been caused merely by superannuation. Long-lived forms belong for the most part to persistent types whose ränge of species is limited. Their reproductive functions have declined, and like an individual in its senescence, they evince the Symptoms of decrepitude and old age. Darwin attributes the extinction of less well-adapted organisms to the struggle for existence ; but since, according to the theory of natural selection, new species arise only with extreme slowness by means of the gradual accumulation of useful variations, and since in like manner their less successful competitors are only very gradually crowded out, we should expect to find in the rocks, supposing that the paleontological record were in any degree perfect, all manner of extinct intermediate forms, and we should be able, at least for those groups especially liable to conservation, to build up complete ancestral trees. But as Observation shows, not only do most plants and animals now living in a wild state adhere to their peculiar characteristics with great tenacity, exhibiting barely appreciable changes even in the course of hundreds or thousands of years, but, furthermore, fossil species remain within the limits of a single geological period fairly constant. With the beginning of a new epoch or period, however, which is usually indicated in the section by lithologic changes, a greater or less number of species either 16 ELEMENTS OF PALEONTOLOGY entirely disappears, or is replaced by closely related, but at the same time more or less different forms. Obviously, therefore, there have been periods when the process of transformation and the weeding out of organisms were greatly accelerated, and following upon these reconstructive periods long intervals of repose have ensued, during which intervals species have retained their characteristic forms with but little Variation. The fact that evolution has advanced by occasional bounds or leaps Stands, however, in nowise contra- dictory to the theory of descent. The whole animate Community at any point on the earth's surface rests normally in a State of equilibrium, the balance being maintained by the com- bined activity of all ranks and members of society. For the preservation of this balance nature practises a most rigid domestic economy. Every plant depends upon particular conditions of soil, food, temperature, moisture and other requisites for its support ; and these conditions govern its distribution and increase in the last degree. Every plant controls the destiny of all animals subsisting upon it ; their numbers multiply with its increase, and wane with its decrease. The fate of these creatures determines that of their natural enemies, who stand in similar relationships to still remoter circles ; and hence no form can overstride the bounds set for it by the general balance without disturbing the whole general System of economy. Let the flora or fauna of a given region become altered by the extinction of a number of species, or by the introduction of new and more powerful competitors, the balance is immediatöly upset. In the first instance vacant places must be filled up, and in the second, room must be made for the newcomers at the expense of the settled Community. Thus, wherever climatal, orographic, or other changes are instrumental in bringing about the extermination of large numbers of plants and animals during the lapse of a geological period, a State of inequi- librium must necessarily result. But thereupon the struggle for existence is waged with unwonted severity among the survivors, until finally a readjustment is established, and a pause in the formation of new species ensues. The whole course of evolution in the organic world during past geological periods indicates not only definite progression in all branches of the animal and vegetable kingdoms up to their present State, but also a more perfected specialisation. Granting that the theory of descent is true, and that all organisms have developed from a single primitive cell, or from a few primitive ground-types, then every new growth and differentiation must stand for im- provement and progress, leading gradually to the development of more or less highly specialised organs, and to a division of labour in their physiological functions ; the higher the degree in which this is manifested, and the more conformably to apparent purpose and utility that each organ fulfils its functions, the more perfect is the organism, as we conventionally term it. Evolution in the organic world has not advanced in a simple, straightforward direction, but its course has been exceedingly complicated and circuitous. The biological Systems, accordingly, do not suggest to us the similitude of a ladder with its numerous rounds, but rather that of an enormously ramifying tree, whose topmost twigs represent the youngest, and, on the whole, the most perfect forms of every brauch. The root, trunk, and a goodly portion of the Upper limbs lie buried in the earth ; and only the ultimate green shoots, the last and most highly difFerentiated members of long ancestral lines, blossom forth in the world of to-day. Phylum I. PROTOZOA. Protozoa are unicellular organisms with bodies consisting of sarcode (proto- plasm), usually very minute, frequently microscopic in size, and without difFerentiated tissues or organs, They are water-inhabitants, take in nourish- ing matter either at any point on the periphery of the body whatsoever, or through a so-called moiith (cytostome), and reject the undigested portions either' from any part of the body whatsoever, or from a definite point called the anal aperture (cijiopijge). The contractile sarcode almost invariably contains one or more nuclei, and exhibits considerable diversity of structure and difFerentia- tion. Locomotion is accomplished by means of vibratile cilia, flagella, pseudo- podia or irregulär processes of the periphery. Reproduction takes place by means of budding or self-division, which latter process is often preceded by a temporary conjugation of two individuals. Protozoa are divided into four classes, only the first-named of which is known to occur in the fossil State : Sarcodina, Flagellata, Infusoria and Gregarina. Olassl. SARCODINA. Protozoa with or without a test, having in fully developed individuals well charaderised pseudopodia, either digitale, reticulate or radiale, with or without axial filaments. Subclass 1. RHIZOPODA. Sarcodina either naked or with a definite test, the pseudopodia either lohose or reticulate ; the adult form is amoeboid. Order 1. AMOEBIDA. The animals constituting this order do not occur as fossils. There are found, however, in chalk and many marine limestones minute calcareous bodies resembling coccoliths, such as are present in vast quantities in deep-sea ooze of existing oceans.^ ^ To the Amoebida were foniierly assigned by Huxley and Haeckel the so-called Bathybius, a reticulated colloidal snbstance coniposed of anastomosing Strands, occuiring at great depths in the Atlantic Ocean. Sir Wyville Thomson and Moebins regarded it as aprecipitate of calcium siilphate, intermingled with decomposed organic matter. In deep-sea ooze, which consists chietly of linie carbouate, as well as in Bathybius, great quantities of minute calcareous bodies of various shapes are found, such as also occur as an essential constituent of chalk, marls and most marine lime- stones belonging to older geological periods (cf. C. W. Gümbel, Neues Jahrbuch für Mineralogie, 1870, p. 7f>3). Ehrenberg termed these bodies morpholites, and regarded them as inorganic in VOL. I 17 C 18 PROTOZOA— RHIZOPODA phylüm i Order 2. FORAMINIFBRA d Orbigny. Bhiso^oda usuaüy wUh a lest wJUch is typkaüy cakareous hut may be säkeous or agghdinaled; consisting of one ar mare Chambers ; pseudopodia räiculaU. The Foraminifera are for the most pari minnte animals varring in size from a fraction of a mülimetre to several millimetres in length, but may deTelop a t^t several inches across ; these, however, are rare exceptions. A few species occur in fresh or brackisli water, but the great majoritj live in the ocean. They are found at all depths, but are most frequent at moderate depths in the ocean basins, where they form characteristic deposits — the so-called "globigerina ooze." In the vicinity of tropical coral islands many species occur in great abundance. The animal itself is a single-celled form with one or many nuclei, as will be later explained The test, in many cases at least^ is really an internal structure, as the thin film of protoplasm which covere it in the perforate forms, and probably in others, is capable of secreting the material of the test to repair fractures and the like. nature. Haxley (Journal Jücrosoop. SdaK«, 16&S, VUI. 2^o. Cj, oJmI Haöck^ ^JcZi^iäeiie Zeit- schrift, 1870^ V. 3, p. 18) regarded thon st fiist as poitions of Baikgtiu*, and des%iiated them coecolühs (Fig. 1). The simple^ dialdike varieties, eonTex ingle nucleus is found in about numerically the middle chamber. In the process of adding a new chamber a portion of the protoplasm is protruded 20 PROTOZOA— RHIZOPODA phylum i from the aperture and a new Chamber wall then formed about it. In some cases a complete wall is formed with each newly added Chamber, but in others the adjacent parts of previous Chambers form the inner walls of the new^ Chamber, and new walls are formed only on the free i)arts of the protoplasmic mass. In the open tubulär test, such as Astrorhiza or Hyperammina, increase in the protoplasmic body is accompanied by addition of material at the open end of the tube and an increase in size results. In single-chambered types, such as Lagena, the manner of increase in size is problematical, if there be any at all. In such forms the entire test may be made in its completed form at once after division, as is the case in certain of the fresh-water Rhizopods. In the tests having more than a single Chamber the apertures of the first- formed Chambers become internal as a rule, and a complexity of relations to the outside medium is thus brought about. One of the simplest arrangements of the Chambers is a linear series. Such an arrangement is seen in Ileophax and Hormosina. Another very common plan of arrangement is a planospiral, as in Ammodiscus. This may be varied by having the revolving line in a spire and then the whole test becomes trochoid, as in Trochammina. Another common arrangement is a biserial one, the Chambers being on opposite sides of the axis, as in Textularia. These four plans or some modification of them are the characteristic arrangements for the Chambers in most of the secreted tests. Oftentimes more than one plan of arrangement enters into the forma- tion of the test. Dimorphism was used for this, but that term has been used elsewhere with a very different meaning. As here viewed, this life-history with several distinct methods of growth has a deeper significance than has usually been attached to it. It seems to have a definite phylogenetic bearing in each particalar group. The term " dimorphism " would hardly cover the case in some genera, where eight or more distinct stages may be made out, each with its characteristic form of Chamber, yet all appearing successively in a Single test. The number of Chambers in the complex tests varies from a few to a great many. Where the size of the test becomes considerable and the Chambers correspondingly large, the Chamber is often divided up in various ways into chamberlets, as in Orbitolites. In such cases the adjoining chamberlets are usually in free communication with one another. The walls of the chamber- lets give additional strength in many forms in which they are developed. Another characteristic modification in some genera is the development of labyrinthic structures in the interior of the Chambers. Such structures are seen in Oyclammina, Haplostiche, Fabularia, etc. In general, it seems tq be a mark of the culmination of certain lines in development, and many of the genera which developed such labyrinthic structures are now extinct. From the appearance of a series of such tests of one species at different stages in development, it would seem as though this labyrinthic condition was developed as a secondary growth in the Chamber. One of its uses may be to give added strength to the test, but this does not always seem to be the case, for it may occur in tests which are characterised by thick walls. The aperture in a given species seems to be rather constant when the development is understood. Much has been written upon this subject ; apertural characters have been used by some authors as a basis for systematic work, and discarded by others as very variable. In a few specimens it may seem at first sight as though the apertural characters were very variable, but ORDER II FORAMINIFERA 21 with a large series showing different stages in development another phase of the matter is presented. In certain cases there is a very decided change in the condition of the aperture, but these changes appear at different stages in the life-history, and all may be seen by cutting back a single full-grown individual. In general, it has seemed from recent studies that apertural characters, when studied in large series, are a rather dependable set for systematic work, and this is true in the Miliolidae and Lagenidae especially. In many species teeth of various sorts are developed in the aperture, and these teeth are subject to various modifications. It can be demonstrated that these modifications occur in a definite sequence, and that this sequence is important from a phylogenetic point of view. In a considerable number of genera a definite tubulär neck is developed, with the aperture at its end. This neck is seen in many genera in a great many modifications, and in Lagena the tube may be inverted and be directed into the chamber of the test. It is obvious that a very long slit-like aperture may be a source of weak- ness to the test, especially when it is at the edge of a thin chamber. Usually in such cases, as in Orhitolites, the animal changes its aperture from a single one in each chamber to a considerable number. This is often coincident with the development of chamberlets, but not invariably so, for multiple apertures occur in PeneropUs where there are no chamberlets. Many of the tests of the Foraminifera are beautifully ornamented. Eaised costae, striations, knobs, spines and punctate areas form the main types of orna- mentation. Several of these or combinations of them may occur in a single species, the form of the ornamentation often changing as the Chambers of the test are developed. Certain of the simpler forms of ornamentation may occur as parallelisms in widely separated groups. As a rule, the proloculum'and early Chambers are smooth and unornamented, but there are certain exceptions, as in Nodosaria, for example, where in some species ornamentation may occur on the first chamber. In specialised genera it is not uncommon to find certain of the species with the early portion of the test ornamented, but the last- formed Chambers with a loss of ornamentation and a consequent development of smooth Chambers. On the other band, there may be a thickening of the test from without and the covering of the Chambers already formed with a secondary growth, often spinöse. Such a condition is seen in some species of Bulimina. Ordinarily the different parts of the test are connected with one another by the previous apertures, but in some cases, notably in Polystomella, there is a secondary canal System which is very complex and runs to all the parts. This has been worked out by Carpenter and others in detail. For many of the Foraminifera two distinct phases have been discovered. One of these, the microspheric form, has a proloculum or first chamber of much smaller size than the other — the megalospheric form. These two forms are to be looked for in all species. The microspheric form (Fig. 4, B) has a number of nuclei, often a larger number than there are Chambers, scattered irregularly through the protoplasm of the body. There seems to be a rather definite relation between the size of the nuclei and the size of the chamber in which they occur, the larger nuclei being in the larger Chambers and the reverse. Apparently these nuclei simply divide in their reproduction during the growth of the test. 22 PROTOZOA— RHIZOPODA rilYLUM I When the animal attains its adult stage tbere is a great increase in the number of pseudopodia, and the entire protoplasm either leaves the test and accumulates aboiit the exterior or is drawn into the outer Chambers. Finally, each nucleus gathers a mass of protoplasm about itself and secretes the proloculum of a new test. This newly formed proloculum is of the larger type and is the first Chamber of the megalospheric form, instead of being of the same size as that of the raicrospheric parent from which it was derived. The megalospheric form (Fig. 4, A) differs from the microspheric form This does not divide, but moves along as new Fig. 4. Biloculinabradyl Schlumb. Receiit ; Bay ofBiscay. A, Small formwith megasphere. B, Large form with microsphere. is/^ (after Schlumberger). in having a single nucleus. Chambers are added, keeping in about the middle Chamber nu- merically. Nucleoli appear in increasing numbers as the growth continues, and finally the whole nucleus breaks down and a great number of minutenuclei appear. These draw aboüt themselves por- tions of the protoplas- mic mass and then di- vide by mitotic division. Finally, the mass leaves the test in the form of zoospores. These are then supposed to con- jugate and to give rise to the small proloculum of the microspheric form, thus completing the life cycle, although the actual process of conjugation has not de- finitelybeen observed in this group. The empty tests left behirid must form a large propor- tion of the dredged Foraminifera. The two forms may be dis- tinguished by the size of the proloculum and, when sufficiently known, by other characters as well Fig. 5. Deep-sea ooze iiiagiiified 700 diameters. a, Bathyhius with Coccolitlis : 1>, Individual Discoliths and Cyatholiths ; c, Coccospheres ; d, Globigerina ; r, Glohifierina with bursted test ; /, Texfularia ; g, g', Radiolaria ; h, i, Diatoms ; Ic, l, Sponge spicules ; m, Mineral fragment. ORDER II FORAMINIFERA 23 The microspheric form is thus the result of a conjugation or sexual process, while the megalospheric form is the result of simple division or an asexual process. As a rule the megalospheric form is by far the most common, and in many species the microspheric form is very rare, or even as yet unknown. The microspheric form, while it Starts as a smaller individual, in most cases attains a much larger size than the megalospheric, as might be suspected from the nature of the reproductive processes by which it is formed. In species where there are definite stages in development it is usually the microspheric form which repeats these most fully, these stages being reduced or entirely skipped in the megalospheric form of the species. In some cases the megalospheric form may give rise to a group of megalo- spheric young instead of to zoospores. On the whole, the life cycle agrees well with the alternation of generations as seen in certain other groups of animals. Fio. 6. Specimen of prepared White Clialk from Meudon, as seen in transmitted light imder power of 300 diameters, showing Texhi- laria, Globigerina, and Rotalia. Fio. 7. Thin slice of Plänerkalk from Bohemia, viewed in transmitted light under power of 50 diameters, showing sections of Nodosaria, Rotalia, Frondicularia, and numerous iso- lated Globigerina Chambers. The vast majority of Foraminifera are marine in habit. They occur in shallow water bordering the coasts, sometimes attached to algae, sometimes creeping on the bottom. A few genera are extraordinarily abundant in the open sea, being found at difFerent depths as free-swimming forms, and also on the floor of the ocean. Enormous quantities of their remains are spread over vast tracts of the sea-bottom, and down to a depth of 2300 fathoms they remain an essential constituent of the deep-sea ooze. This is a finely divided agglomeration of decomposed calcareous substances, such as the shells of mollusks, corals, bryozoans, coccoliths, radiolarians, diatoms, sponges and Foraminifera. Of the latter, certain genera are remarkable for their extra- ordinary abundance (Globigerina, Orbulina, Pulvinulina, Biloculina) (Fig. 5). In the Atlantic and Pacific Oceans Globigerina ooze is the prevailing deep- sea deposit; in the North Sea, along the coast of Norway, Biloculina ooze. Numerous limestones and marls of older geological periods exhibit great 24 PROTOZOA— EHIZOPODA phylüm i similarity in structure and chemical composition to the now forming deep-sea oozes. White Chalk (Fig. 6) is clearly a variety of abyssal ooze, from which siliceous constituents have become segregated out, and in which Textularia predominate instead of Glohigerina. Certain of the Eocene limestones of the Paris basin are composed almost exclusively of the tests of Müiolidae, while others are made up of Älveolinae and Nummulites. During the Carboniferous period the chief role as rock-building organisms was played by Fusulina. Many dense, apparently homogeneous, or even semi-crystalline limestones of various ages, when examined microscopically in thin sections, are seen to be composed in large part of Foraminifera and other organic bodies (Fig. 7). Fossil Foraminifera are best preserved, being usually detachable from the matrix, and at the same time occur most abundantly, in unconsolidated marls and clays which are interbedded with calcareous strata, or in limestones of a chalky or earthy character. The tests of Foraminifera were first discovered by Janus Plancus, in 1730, on the beach of Rimini, and in the following year they were found by Beccari in the Pliocene of Bologna. They were long considered to be shells of mollusks, and were described by Breyn, Soldani, Fichtel, d'Orbigny and others as Cephalopoda foraminifera, in distinction from Cephalopoda sipfionifera. Dujardin, in 1835, was the first to recognise their true character as belonging to the Rhizopoda. Family 1. Gromidae. Test chitinous with an aperture at one or both ends for the pseudopodia. The animals belonging to this family are mostly fresh-water species and their occurrence as fossils is unknown. Family 2. Astrorhizidae Brady. Test composed of agglutinated material for the most part, occasionally with a chitinous inner layer, consisting of a Chamber with several openings or a tubulär test open at both ends; or in certain forms of a closed Chamber with a single aperture. Throughout the family the test is not divided into a series of Chambers. Becent and very abiindant at depth. Fossil in Paleozoic and later formations. Subfamily A. Astrorhizinae Brady. Test consisting usually of a tube open at both ends or with several tubes entering a central Chamber ; in some species with the tube branching. The genera Astrorhiza, Bhabdammina, Marsipella, Bathysiphon and Rhizammina make up this subfamily. Apparently fossil since the Upper Jurassic and common in Recent. Subfamily B. Saccammininae Brady. Test consisting of a single Chamber, or group of superficially attached Chambers. The walls made up for the most part of agglutinated material ; apertures sometirnes numerous but usually single ; tests free or attached. K HÄRDER II FORAMINIFERA 25 ^^H Saccammina Sars. (Fig. 8). Shell thick, with labyriiithiform interior ; ^^■pherical, pear-shaped or fusiform, with tubulär prolongations at one or both ^^pnds ; sometimes united together in chains. Ordovician (Ayrshire), Devonian (Canada), Carboriiferous and Recent. Entire strata of Carboniferous rock near Elfhills, Northumberland, are built up by aS". carteri Brady. Large-sized species of Astrorhiza, Fsammosphaera, Saccariimina, Hyperammina, and Rhabdammina are described by Häusler from the Upper Jurassic (Trans- versarius beds) of Switzerland. Thurammina Brady. Test free, monothalamous, irregularly spheroidal, usually with excrescences or spiny processes. Upper Jurassic and Recent. Siibfamily C. Hyperammininae. Test consisting of a globular proloculum and a more or less elongated, some- times hranching portion, but not divided iiito Chambers; free or attached wall of various agglutinated materials. The genera Hyperammina, Saccorhim, Tolypammina, Ammolagena, Jaculella and Sagenina make up this subfamily. Some of these occur as fossils. Subfamily D. Ammodiscinae Ciishman. Test composed of a globular proloculum and long undivided tube, closely coiled, either planospirally- or in changing planes or to form a spiral lest ; wall of fine sand with much cement. Ammodiscus Reuss. Test free, composed of a proloculum and long coiled tubulär chamber. Carboniferous to Recent. Family 3. Lituolidae Brady. Test composed of agglutinated material foi' the most part ; consisting of two or more Chambers ; arranged in a linear, coiled or irregulär series ; apertures usually one to each chamber, but sometimes more. The tests included in this family all have the wall composed of ag- glutinated material with a varying amount of cement in the different genera. Throughout the family as here used the tests are composed of two or more Chambers and a definite proloculum is apparent. Usually the tests are composed of a series of Chambers. Subfamily A. Aschemonellinae Cushmaii. Test composed of agglutinated material, divided irregularly into Chambers mthout a defiyiite ptlan of arrangement. Subfamily B. ReophäCinae Cushmaii. Test of agglutinated material, sand grains, sponge spicides, etc., with a varying amount of cement, Chambers in a linear series, aperture single at the distal end of the last-formed chamber. Reophax Montfort. Test free, composed of a lineal series of Chambers, 26 PROTOZOA— EHIZOPODA PHYLUM I joined end to end in iiearly a straight line, curved but not coiled, wall coarsely arenaceous, Chambers undivided, aperture simple and terminal. Carboniferous to Recent. Haplostiche Reuss (Fig. 9). Test similar to Beophax but the Chambers divided into labyrinthic cavities, aperture in adult made up of several pores or dendritic. Jurassic to Recent. Subfamily C. Trochammininae Brady. Test composed of several Chambers, either in a planospiral coli, trochoid or other- wise arranged ; wall composed of sand grains of varying degrees of coarseness cemented with a calcareous or ferruginous cement ; free or attached. Trochamminoides Cushm. (Trochammina Reuss, pars) (Fig. 10). Test free, composed of several coils, each constricted into a number of Chamber -like portions with large openings between ; wall of fine sand and yellowish-brown A, Saccammina carteri Brady. Carboniferous Lime- stone ; Elf hüls, Northum- berlaiid. 1/2. B, Fractured test, filled with interior cal- cite. 10/j (after Brady). Haplostiche horrida Schwa- ger. Upper Jurassic (Im- pressa clay) ; Gruibingen, Württemberg. Fig. 10. Trochammina pro- tcws Karrer. Vienna sandston e (Senonian) ; Hütteldorf, near Vienna. Fig. ]]. Lituola (Haplo- phragmium) irregu- laris Roemer. Sca- phiten - Pläner ; Kröndorf, Boheniia. Fig. 12. Placopsiliria r od rata Quenst. sp. Upper Jur- assic (Inipressa clay) ; Reichen- bach, Württem- berg. cement, aperture simple at the end of the last-formed Chamber. Lias to Recent. Ammobaculites Cushm. {Haplophragmium Reuss, pars) (Fig. 11). Test free, chambered, early portion close-coiled in one plane, later portion uncoiled and made up of a more or less linear series of Chambers ; wall coarsely arenaceous, fairly thick ; aperture single, at the centre of the terminal face of the uncoiled portion, but in the coiled portion at the base of the apertural face.. Carboni- ferous to Recent, particularly abundant in the Jurassic and Cretaceous. Placopsilina d'Orb. (Fig. 12). Test rugose, arenaceous, attached, and divided into pyriform or spherical Chambers, which are joined in chains or are irregularly united. Lias to Recent. Subfamily D. Neüsininae Cuslimaii. Test arenaceous with som,e chitin, broad and flattened, of many Chambers, early portion coiled, later Chambers broad and spreading ; sides with elbngated chitinous ßaments. Here is placed the single recent genus Neusina Goes. Fig. 13. Orhitolina concava Lam. ORDER II FORAMINIFERA 27 Subfaniily E. Orbitolininae. Test siliceoiis, imperforafe, crateriform and. composed of concentric annuli which iire partitioned off into numerous Chambers. Orhitolina Lam. (Fig. 13). Test composed of agglu- tinated sandy particles ; bowl - shaped to depressed conical ; upper surface convex, lower slightly concave ; externally smooth or with concentric bands. Test composed of multilocular rings, the Chambers com- municating with one another on all sides by pores. The outer portion of each Chamber is subdivided by two secondary partitions disposed at rieht anerles to cenomanian; Urscheiau, , ., -XT 1 i.'ii T 1 TT Bavarian Alps, a, Inferior each other. Very abundant in the Lower and Upper surface; b, superior sur- Cretaceous. 0. lenticularis and 0. concava Lam. (eniaIrg'edV''''''' '''*'''" Family 4. Textulariidae Schnitze, Test either arenaceous or calcareous, perforate, the Chambers usually nnmerous, essentially biserial or triserial, or in some genera spirally arroMged. The family Textulariidae is apparently the most primitive, after the Lituolidae. A number of the genera are wholly or in part composed of species with arenaceous tests, which is in itself a primitive character in the group. In many species both the microspheric and megalospheric forms are known. In the microspheric form, which repeats most completely the phylogenetic characters, a coiled early development succeeding the proloculum is commonly found. This stage may be compared to the entire development of such a genus as Haplophragmoides in the Lituolidae. Subfamily A. Spiroplectinae Cushman. Test either coarsely arenaceous or calcareous, or even hyaline, the early Chambers following the proloculum closely coiled, the later Chambers biserial, occasionally tending to become uniserial in the last developed Chambers. This subfamily includes the single genus Spiroplecta Ehrb., which in its developmental stages connects the Textulariidae with the Lituolidae. Its de- velopment is primitive in that the stages are seen in both the microspheric and megalospheric forms of the species, and are of comparatively long duration. Cretaceous and post-Tertiary. Subfamily B. Textülarinae Brady. Test typically biserial, early portion in microspheric form offen with a few coiled Chambers, followed by biserial ones, later Chambers variously modified in diff'erent genera, uniserial, broadly extended, etc. Wall either arenaceous or calcareous and hyaline, perforate ; aperture single, or in a few cases, many present in a single Chamber. Textidaria Defr. (Fig. 14, A). Test usually elongated, straight, tapering, or turbinated. Chambers biserial, altcrnating and communicating with each 28 PßOTOZOA— KHIZOPODA PHYLUM I other by means of slit-like apertures. Carboniferous to Recent. Extremely abundant in the White Chalk. Fig. 14. A, Tbxtularia globifera Reuss. Upper Cretaceous (Senonian) ; Pattenauer Stollen, near Traunstein, Bavaria. B, Bolivina incrassata Reuss. Upper Cretaceous ; Götzreuther Graben, near Siegsdorf, Bavaria. C, Flecanium gihhosum d'Orb. Pliocene ; Sienna, Italy, D, Grammostomum {VidvuUna) gramen d'Orb. Recent; Cuba. E, Gaudryina rugosa d'Orb. Upper Cretaceous. Götzreuther Graben, near Siegsdorf. F, ClavuUna communis d'Orb. Miocene ; Baden, near Vienna. Bolivina d'Orb. (Fig. 14, B). Test biserial throughout, aperture elongate, Fig. 15. ^.r;t' tf^mT ^'''X'^"'"' 'i'O^.b- Miocene (Leitbakalk) ; Nussdorf, near Vienna. B, Bulimina pupoicks d'Orb. Snal section iwXTM'^r\*'''n ^fv-^"""™'' ^*^"'"- .Carboniferous Limestone Dugno, Rus^sia. Longil SeTsi^KenerBaTetnear^;^^^^^^^^^ ^^^^^^°' ^"^^^^^ "''^ (^^teV Möller). ' ., ^. J^^e^^in^^^^: usually wider at one end, hyaline in young, thickened with age. Cretaceous to Recent. ORDER Ti FOEAMINIFERA 29 Climacammina Brady (Crihrostomum Möller), (Fig. 15, C, D). Test arenaceous with calcareous basis. Chambers biserial, rectilinear. Oral aper- ture porous. Abuiidaiit in Carboniferons Limestoiie (cf. Bigenerina d'Orb.). Siibfaiuily C. Verneuilininae Cußliman. Test cd first triserial, later biserial or even uniserial in some genera. Gaudryina d'Orb. (Fig. 14, E). Test free, early portion triserial, later Chambers arranged biserially, wall usually arenaceous. Clavulina d'Orb. (Fig. 14, F). Test at first triserial, latest developed portion uniserial. Eocene to Recent. Valvulina d'Orb. (Fig. 15, E). Test arenaceous with calcareous basis. Chambers in triple series arranged in screw-like spiral. Qarboniferous to Recent. Tetrataxis Ehrbg. (Fig. 15, F). Test calcareous, conical. Alternating Chambers arranged in a turbinate spire. Carboniferous Limestone. Siibfamily D. Bulimininae Brady. Test composed of Chambers in an elongate spiral, aperture elongate, loop-shaped, usually oblique, lest calcareous, hyaline in young. Bulimina d'Orb. (Fig. 15, J, B). Test calcareous, the alternating Chambers arranged in an elongated spire. Triassic to Recent. Subfamily E. Cassidülininae Brady. Test with the Chambers biserial but combined with a spiral or volute arrangement making a complex lest. EhrenberginaRenss (Fig. 15, G). Test calcareous, the alternating biserial Segments either completely or only partially coiled. Tertiary and Recent. Cassidulina d'Orb. Tertiary and Recent. Family 5. Chilostoniellidae. Test calcareous, finely perforate, composed of numerous Chambers, following each other from the same end of the long axis, or alternately from the two ends, or in cycles of three. The genera Ellipsoidina Seg., Chilostomella and Allomorphina Reuss compose this family. Cretaceous to Recent. Family 6. Lagenidae Carpenter. Test calcareous, vitreous, finely perforated, one or more Chambers placed in a straight line, coiled or variously arranged. < Subfamily A. Lageninae Brady. Test monothalamous, flask-like. Lagena Walker (Fig. 16, A). Test single-chambered, spherical, ovate or flask-shaped, with terminal oral aperture. Silurian to Recent. 30 PROTOZOA— RHIZOPODA PHYLUM I Subfamily ß. Nodosarinae Brady. Test either coiled or uniserial, or a modification of one or the other. Nodosaria Lam. (Fig. 1 6, B). Test rod-shaped ; Chambers arranged in a linear series and set off from one another by constrictions ; oral aperture round, terminal. Abundant and widelv distributed from Silurian to Recent. Fig. 16. A, Lagena semistriata Williamson. Antwerp Crag (Pliocene) ; Antwerp. B, Nodosaria spinicosta d'Orb. Tegel (Miocene); Baden, near Vienna. (', Dentalina elegans d'Orh. Same locality. D^Cristellariarotulata'La.m. Scapliiten- Pläner (Turonian) ; Boheniia. E, Va/jinulina recta Reuss. Neoconiian ; Salzgitter, Hanover. F, LinguUna costata d'Orb. Tegel (Miocene) ; Baden, near Vienna. Dentalina d'Orb, (Fig. 16, 6'). Like the preceding, but test slightly arcuate. Carboniferous to Recent. LinguUna d'Orb. (Fig. 1 6, F). Test rectilinear, compressed ; segments regularly attached; aperture terminal, slit-like. Trias to Recent. Glandulina d'Orb. (Fig. 17, A). Test abbreviate, ovate ; segments united in rectilinear series, half embracing one another. General aperture round, terminal, tubiform. Trias to Recent. Faginulina d'Orb. (Fig. 16, E). Test rectilinear, laterally compressed; segments flattened, with obliquely directed septa. Trias to Recent. Margin ulina d'Orb. Early portion arched or helicoid, later segments recti- linear. Terminal aperture slit - like. Trias to Recent. Cristellaria Lam. (Fig. 16, D). Test regularly planospiral, with convolutions completely envelop- ing one another. Ter- minal aperture round. Trias to Recent. Frondicularia Defr. (Fig. 1 7, D). Test extremely compressed and foliately expanded in a single plane; Chambers r^flexed and laterally embracing one another. Terminal aperture round. Trias to Recent. Ä, Glandulina inflata Boruem. Septarienthon (Oligocene) ; Hennsdorf. B, Polymorphina inflata Williamson. Recent ; German Ocean. C, Dimorphina sp. Pliocene ; Sienna, Italy. D, Frondicularia goldfussi Reuss. Scaphiten- Pläner ; Dülmen, Westphalia. E, Uvigerina p'ygmaea d'Orb. Tegel (Miocene) ; Baden, near Vienna. ORDER II FORAMINIFERA 31 Subtaiiiily (J. Polymorphininae Brady. Test composed of Chambers arranged spirally or irregularhj ahout the long axis ; aperture usually radiale. Polymorphina d'Orb. (Fig. 17, B). Segments irregularly helicoid, or arranged biserially, more or less enveloping one another and variable in shape. Terminal aperture round. Trias to Recent. Dimorphina d'Orb. (Fig. 17, C). Early Chambers irregularly or triserially arranged, later ones foUowing in rectilinear fashion. Cretaceous to Recent. Subfaniily D. Uvigerininae Cnshman. Test composed of Chambers arranged triserially ahout the long axis ; aperture usually simple, with a deficite neck and a phialine lip. Uvigerina d'Orb. (Fig. 17, E). Segments dissimilar, disposed in triple series, and spirally wound like a gastropod shell. Eocene to Recent. Subfamily E. Ramulininae Brady. Test composed of Chambers with long tuhulariform tubes. Ramulina Rupert Jones. Test branching, consisting of rounded Chambers joined by stolon-like tubes. Recent, and possibly also represented in the Cretaceous. Family 7. Globigerinidae Carpenter. Test free, calcareous, perforated by coarse tubules ; monothalamous or poly- thalamous ; Chambers globular, either irregularly disposed or imperfectly spiral. Of the tvvo principal genera belonging to this family, Orbulina d'Orb. (Fig. 18, A) is unilocular, and GloUgerina d'Orb. (Fig. 18, C) is multilocular. The individual Chambers usually open into a common central canal. In both genera the test is often covered with extremely delicate calcareous spines, which, however, are very easily broken off, and are never preserved intact in the fossil State. Both these genera are excessively abundant in existing oceans (Globigerina ooze) ; they occur sparingly in the Trias and throughout the Mesozoic, first becoming important during the late Tertiary. Sphaeroidina d'Orb. (Fig. 18, B). Characters few, so coiled as to form a nearly globular test ; aperture with a valvulär lip. Cretaceous to Recent. Fig. 18. A, Orhulina universa Lam. Pliocene ; Sieniia, Italy, B, Sphaeroidina austriaca d'Orb. Miocene Tegel ; Baden, near Vienna, C, Globigerina con- glomerata Schwager. Pliocene ; Kar Nikobar Island, a, Inferior siirface ; h, Superior surface ; c, Portion of periphery ; d, Transverse section enlarged. 32 PROTOZOA— EHIZOPODA PHYLUM I Family 8. Rotalidae Carpenter. Test calcareous, perforate, free or adherent, typicalhj spiral at hast in the young. Siibfamily A. Spirillininae Brady. Test a flat spiral, without divisions, free or attached. Spirillina Ehrenberg. Test a planospiral undivided tube, free or attached. Miocene to Recent. Subfamily B. Eotalinae Carpenter. Test calcareous, rarely arenaceous or siliceous, finely or coarsely perforated, frequently with intermediate skeleton, free or adherent, turbinate or discoidal in contour. Segments usually arranged in an elongated spire, although in some forms irregularly Discorhina Parker and Jones (Fig. 19, A, B). Test coarsely perforated, Fig. 19. Ä, Discorhina (Asterigerina) planorbis d'Orb. Miocene (Leithakalk) ; Nussdorf, near Vieima. B, Discorhina sp. Recent. a, Under side ; h, Upper side ; c, Lateral view ; (/, Median section. C, Planorhulina mediter- ranensis d'Orb. Recent ; Mediterranean. a, Inferior surface ; h, Superior surface ; c, Transverse section. turbinoid ; lower surface broad and flat ; umbilicus often filled with deposit of intermediate skeleton. Cretaceous to Recent. Fig. 20. A, Rotalia heccari Lin. Pliocene ; Sienna, Italy. B, Pulvinulina partsiM d'Orb. Miocene (Tegel); Baden, near Vienna. C, Endothyra panderi Möller. Carboniferous Limestone; Russia. 20/^, j)^ Endothyra parva Möller. Lower Carboniferous ; Russia. Longitudinal section. loo/^. Planorhulina Parker and Jones (Fig. 1 9 G). Test coarsely perforated, com- planate, usually attached, upper and lower surfaces dissimilar : early segments 'UDER II FORAMINIFERA 33 arranged in a depressed spire, subsequeiitly becoming cyclical. Lias to Recent. Various subgeiiera, named by d'Orbigny Truncatulina, Anomalina, Planulina, etc., are based upon slight modifications in form. Rotalia Lam. (Fig. 20 Ä). Test finely perforated, with segments in turbinoid spire. Septa composed of two slightly separated lamellae, with anastomosing canals occupying the intermediate space. Base often thickened by supplemental skeleton. (?) Silurian. Upper Jura to Recent. PulvinuUna Parker and Jones (Fig. 20 B). Rotaliform, but septa simple without being per- forated by a canal System. Lower Lias to Recent. Endothyra Phill. (Fig. 20 C, D). Test cal- careous, composed of an external coarsely per- forated and an internal compact layer, the latter finely granulär; segments numerous, coiled in an irregulär spiral, terminal Chamber opening by several apertures. Abundant in Lower Carbon- iferous, and existing at the present day, ac- cording to Brady. Calcarina d'Orb. (Fig. 21). Test discoidal, with dissimilar upper and lower surfaces ; Chambers spirally coiled. Exterior encrusted with a supplemental skeleton which fills up all depressions and forms spinous or spur-like processes traversed by coarse canals. Upper Cretaceous to Recent ; very abundant in Maestricht Chalk. Fig. 21. CalcMrina calcitrapoides Lam. Upper Cretaceous (Tuffkreide) ; Maestricht, Holland. Siibfamily C. Tinoporinae Brady. Test of irregularly massed Chambers, the early anes more or less distinctly spiral in their arrangement, usually without a general aperture. Tinoporus Montf. Patellina Williamson. The Recent genera Carpenteria Gray, Rupertia Jones, etc., are distinguished by their extremely irregulär, coarsely perforated and usually adherent tests, which sometimes attain considerable size and often contain agglutinated, Sandy or various other foreign particles. Thalamopora Roemer, occurring in the Cretaceous, probably also belongs to this subfamily. Family 9, Nummulitidae. Test calcareous, finely tubulated, polythalamous, free, spiral, usually bilaterally symmetrical. Subfamily A. Fusulininae Brady. Test fusiform or siibglobular Chambers extending from pole to pole, each convolu- tion completely covering the preceding whorls. Schwagerina Möller. Test spherical, finely perforated. Primary and secondary septa simple, thin, straight ; secondary chamberlets communicating with the next following principal chamber by means of a basal aperture. Abundant in Lower Carboniferous rocks of Japan, China, Sumatra, North America and Russia. VOL. I D 34 PROTOZOA— RHIZOPODA PHYLÜM I Fusulina Fischer (Fig. 22). Test fusiform, laterally elongated like Alveolina, coarsely perforated. Septa of principal Chambers undulating, and Fig. 22. Ä, Fxisulina cylindrica Fisch. Carboniferous Limestone ; Saranisk, Russia. Natural size. B, C, Same species showing various cross-sections enlarged. D, Enlarged section showing Chambers communicating by means of foramina (a, b). united so as to form secondary chamberlets. Excessively abundant in the Lower Carboniferous of Europe (Russia), Asia and North America. Subfamily B. Polystomellinae Brady. Test Ulaterally symmetrical, nautiloid, the more complex IS8 specimens with a well-developed secondary canal System. Fig. 23. Polystomella crispa Lam. Pliocene ; Sienna, Italy. (Highly magnified.) Polystomella Lamarck (Fig. 23). Test regulär, equi- lateral, nautiloid, final whorl alone visible from the exterior. Jurassic to Recent. Subfamily C. Nummulitinae Brady. Test lens-shaped or flattened, higher forms with complex secondary canal System. Archaediscus Brady. Test lenticular, unsymmetrical, spirally coiled. The Segments irregularly constricted and expanded so as to form Chambers. Septa and canal-system wanting. Lower Carboniferous. Fig. 24. Ämphisteginahaueri d'Orb. Miocene (Leithakalk) ; Nuss- dorf, near Vieniia. a, Exterior Views, enlarged ; h, Natural size ; c, Median section, greatly enlarged ; d, Trans- verse section, greatly en- larged. Fig. 25 Operculina complanata (Ji&st.). Miocene; Bor- deaux, a, Natural size ; h, c, Median and longi- tudinal sections, greatly enlarged. Heterostegina costata d'Orb. Miocene (Leithakalk) ; Nuss- dorf, near Vienna. Amphistegina d'Orb. (Fig. 24). Test lenticular, slightly inequilateral, spirally rolled. Whorls divided into Chambers by numerous single septa in ORDER II FORAMINIFERA 35 which canals are not present; solid wedge-shaped deposit of intermediate Fig. 27. Nummulites cfr. lucctsamts Dfr. Eocene ; Kressenberg, Upper Uavaria. Several times enlarged. a, Marginal cord with canal-system ; ?^ Septal plane with interseptal canal-system ; c, Interior of Chamber ; d, Finely perforate periphery ; e, Small pillars of intermediate skeleton. skeleton near the umbilicus. On one side the volutions completely enclose one another as far as the centre, on the other they overlap only partially by meaiis of alar proloiigations extending in- wards. Chambers communicate with each other by means of a slit along the basis. Miocene to Recent. Particularly abundant in Miocene. Operculina d'Orb. (Fig. 25). Test discoidal, complanate, composed of three to six rapidly expanding spiral whorls, which are polythalamous and non-involute. Septa and marginal cord traversed by a direct canal-system, which gives off numerous branches. Cretaceous to Recent. Particularly abundant in Eocene. Heterostegina d'Orb. (Fig. 26). Like Ojperculina, but with Chambers sub- divided by secondary septa into chamberlets. Tertiary and Recent. Fig; 28; Nummulites (Assi- lina) exponens Eocene Sow. Pyrenees. C 3 C' 1 2 Fig. 29. A 1 ••i, Nummulites (jizchensis Ehrbg. Eocene ; Libyan Desert. Natural size. A -^ Specimen with eroded peri- pheral i)ortion, showing arrangement of septa. B^^, Nximmulites laevigatus Lam. Calcaire Grossier ; Paris. Natural size. B'i, Portion of same enlarged. C^^, Nummulites ramondl Defr. Eocene (Nunimulitic lime- stone); Pyrenees. Natural size. C 3, Enlarged section. Nummulites Lam. (Phacites Blumenb. ; Lenticulites Lam.) (Figs. 27-29). Test 36 PROTOZOA— EHIZOPODA PHYLÜM I symmetrically leiiticular or discoidal, composed of numerous spirally arranged polythalamous volutions, and usually with columnar intermediate skeleton, which forms small excresceiices on the periphery. The septa and marginal cord contain a coarse, anastomosing canal-system, as in Operculina. Primordial Chamber spherical, sometimes large, sometimes exceedingly minute in size. The whorls either merely embrace one another {Ässilina) (Fig. 28), or they completely envelop one another by means of alar prolongations reaching inwards to the centre (Nummulina). The septa are pierced in the median plane by an oblique slit-like aper- ture, and also extend into the saddle- shaped alar prolongations of the Chambers. They are directed in the groups Badiatae and Striatae in straight or slightly curved lines (Figs. 27 and 29, C) ; in the Sinuatae they follow meandering courses (Fig. 29, A) ; and in the Eeticulatae (Fig. 29, B) they form an interlacing network by means A, Nummuiitic limestone with horizontal sections of of connecting processcs. The rami- N distans Fusch Peyrehorade, in the Pyrenees. B, fications of these lateral prOCeSSCS Nummuiitic limestone showing sections of N. lucasanus w « wx i. v^ ^ Defr, zakophane in the Carpathians. (filet cloisomiaire) may be rcadily Seen on fracturing a portion of the test, and are a valuable aid in the determination of species. The oldest Nummulites {N. pristinus Brady) occur very sparsely in the Carboniferous limestone and Upper Jurassic, but are distinguished from the typical later forms by the absence of an interior canal-system in the marginal cord. The typical Nummulites which are so characteristic of the Eocene (Nummuiitic limestone) in Europe, North Africa, Asia and Central America, often build up massive formations. The largest species (JV. gizehensis Ehrbg., N. orUculatus Schaf h.) attain a diameter of 60 mm. ; the smallest species does not exceed 2 mm. ; recent representatives comparatively scarce. Subfamily D. Cycloclypeinae Brady. Test flat with a thicJcened centre, or lens-shaped, consisting of a diso of Chambers arranged in concentric annuli with peripheral thickenings, septa double with inter- septal canals. Orbitoides d'Orb. {Hymenocyclus Bronn; Lycophrys Monti.) (Fig. 31). Test discoidal, with circular or stellate contour, often bent, exterior smooth or with radial Striae, and composed of numerous concentric annuli disposed about a primordial spiral of three to five whorls. The rings are divided by transverse partitions into small rectangular Chambers, and the septa and marginal cord are traversed by canals. Superimposed over the median series of principal Chambers on both sides are several layers of flattened secondary chamberlets, which are likewise disposed in concentric rings. Very abundant in the Eocene, associated with Numm,ulites ; rare in Upper Cretaceous and Miocene. Cycloclypeus Carp. Miocene and Recent. »ER II FORAMINIFERA 37 I ^^B Dawson, Carpenter and various other authors have referred the so-called ^^fozoon occurring in crystalline limestone of the Archaean (Laurentian) period to the Foraminifera ; biit the elaborate investigations of Möbius have shown in, liqn. deL. Fig. 31. A, Orbitoides papyracea Boubee. Eocene (Ferruginous sandstone) ; Kressenberg, Upper Bavaria, (Greatly enlarged). i Median Chambers ; 2 Lateral Chambers ; « Compact pillars of intermediate skeleton. B, Portion of median transverse seetion, highly magnified ; 2 Lateral Chambers with perforate walls ; 4 Canal - System of cyclical marginal cord ; ^Tabules connecting adjacent Chambers. C, Periphery and profile of same, natural size. D, Orbitoides tenella Gümbel. Eocene ; Kressenberg, (Natural size). E, Orbitoides variecostata Gümbel; Eocene ; San Martino, near Verona. (Natural size.) F, Orbitoides ephippium Sow. Eocene ; Kressenberg. (Natural size.) that neither Eozoon nor Archaeosphaerina can be regarded as organic structures, being merely mineral segregations. Family 10. Miliolidae Carpenter. Test of one or more Chambers, calcareous and porcellanouSj sometimes covered with sand, usually imperforate, hut in some forms with the early Chambers distinctly perforate. Subfamily A. Cornuspirinae Ciishman. Test planospiral, usually of a proloculum and long coiled Single Chamber. Cornuspira Schnitze (Fig. 32). Test composed of numer- ous plano-spiral convolutions ; oral aperture simple, terminal; monothalamous. Lias to Recent. Fig. 32, Cornuspira polygyra Reuss, Oligocene ; Hungary. Subfamily B. Nubeculariinae Brady. Test irregulär and asymmetrical, the apertures variously placed. Nubecularia Defrance. Test at first coiled, later tubulär or irregulär ; attached. Liassic to Recent. 38 PEOTOZOA— EHIZOPODA PHYLÜM I an\ Subfamily C. Häuerininae Brady. First-formed part of test Cornuspii'a-like, later Chambers Spiral or otherwise apertures Single. Ophthalmidium Kubier. Early Chambers like Cornuspira, later ones two or more to a convolution, Liassic to Eecent. Ilauerina d'Orbigny. Early Chambers Milioline, later ones planospiral with two or more Chambers to a convolution. Cretaceous to Recent. Subfamily D. Miliolinae Brady. Test at first spiral, then eacli whorl divided typically into two Chambers, laier Chambers more numerous in the whorl or uniserial. Miliola Lam. (Figs. 33, 34). Chambers disposed in coil-shaped loops about a few spirally wound primordial Chambers. Each Chamber in the Fig. 33. A, Biloculina inornata d'Orb. From the Miocene Tegel ; Baden, near Vienna. B, Triloculina gibha d'Orb. Oligocene sand from Astrupp. C, Sjnroloculina hadensii d'Orb. Miocene Tegel ; Baden, near Vienna. D, Quin- qüeloculina saxorum d'Orb. Eocene (Calcaire Grossier) ; Grignon, near Paris. Fig. 34 A, Longitudinal section of Biloculina inornata d'Orb. (enlarged). B, Transverse section of Quinquelomlina saxorum d'Orb. (enlarged). adult forms a half coil. Terminal pseudopodial aperture either curving in the form of a crescent about a tooth-like protection, or branching den- dritically (Lacazina). Forms having all the segments disposed in a single plane, and all externally visible, are grouped together in the genus Spirolocu- lina d'Orbigny ; with all the segments completely enveloping one another, Biloculina d'Orb. ; segments disposed in three or in five dilferent planes, Miliolina AVill. ( = Trilocu- •= ^^ lina and Quinqueloculina d'Orb.). The great variety and profusion of these genera combine to make them some of the most important of the rock- building Foraminifera. Massive beds of Eocene limestone (Paris basin, Pyrenees) are made up of Miliola remains ; at the present day calcareous deposits are being formed by Biloculina in the North Sea west of the coast of Norway. Miliola first makes its appearance in the Trias, and attains its maximum development in the Tertiary and Recent periods. Fig. 85. Fahularia dis- colithes Defr. Eocene (Calcaire Grossier) ; Paris. Fig. 36. Vertebralina mu- cronata d'Orb. Recent ; Medi- terranean. ORDER II FORAMINIFERA 39 Fahularia Defr. (Fig. 35). Like Biloculina, but relatively larger. General aperture cribriform; Chambers not an empty cavity, but filled with porcell- anous or calcareous matter, and perforated by numerous anastomosing canals which are directed parallel with the axis of convolution. Abundant in the Eocene of the Paris basin. Vertebralina d'Orb. (Fig. 36). First-formed portion of test consisting of coil-shaped loops, the segments afterwards becoming joined in rectilinear series. Tertiary and Eecent. Idalina Schlumb. Last -for med Chamber completely enveloping all pre- ceding ones. Cretaceous. Subfamily E. Peneroplinae Brady. Test planospiral or cycUcal, and bilaterally symmetrical ; apertures many. Peneroplis Montf. (Fig. 37). Test discoidal, complanate, polythalamous ; direction of growth primarily spiral, gradually becoming rectilinear, while rapidly increasing in width. Septa perforated by numerous pores. Tertiary and Recent. A ^^'V^T^r<-:W Fio. 37. Peneroplis planatun Montfort. Recent ; Mediterranean. Fig. 38. Orhiculina nummis- 7?mWsd'Orbigny. Plio- cene ; Sienna, Italy. Fig. 39. A, Orbitolites complanata Lam. Eocene (Cal- eaire Grossier); Paris. B, Portion of same en- larged. Orhiculina Lam. (Fig. 38). Test discoidal, first-formed portion spiral, afterwards becoming annular ; polythalamous ; septation regulär, Chambers subdivided ; septa and walls of segments perforate. Tertiary and Recent. Orbitolites Lam. (Fig. 39). Test discoidal, circular in outline, both surfaces slightly concave in the middle, attaining comparatively large size, and com- posed of segments which are arranged concentrically about a few spirally coiled primordial Chambers. Septa radially disposed, and perforated by symmetrically placed pores. In the more complicated forms the principal segments are invested with a superficial multilocular layer, the Chambers of which are also arranged in concentric rings and communicate with the principal segments by means of pores. An important rock-building genus, ranging from the Lias onward. 0. praecwsor and 0. circumvuha Gümbel are Jurassic, 0. macropora d'Orb. Cretaceous, and 0. complanata Lam. Tertiary species. Subfamily F. Alveolininae Br£f9y. Test spiral, elongated in the direction of the axis of coiling ; Chambers divided into secondary chamberlets. ^,^- '~ Aheolina d'Orb. (Borelis Montf.) (Fig. 40). Test> >> .2 1 1 Gromidae Astrorhizidae Lituolidae Textulariidac Chilostomellidae Lagenidae Glohigerinidae Rotalidae Nummulitidae Miliolidae ' 1 ^ 1 [The foregoiiig chapter on Foraminifera has been revised for the preseut work by Dr. Joseph A. Cushnian of the Boston Society of Natural History, Boston, Mass.— Editor.] Orders. RADIOLARIA Müller.^ {Polycystina Ehrenberg.) Marine FJiizopoda emitting fine, filiform, radially direded pseudopodia, with central capsule and extra-capsulum, and usually with delicate siliceous skeleton. The sarcode body of the Radiolarians is differentiated into (1) an inner central sphere or capsule of tough gelatinous-like protoplasm containing one or ^ Literature : Ehrenberg, G., Mikrogeologie, 1854 ; also memoirs on Radiolaria from Barbados, in Al)handl. Akad. Wiss. Berlin, 1872, 1875.— Haeckel, K, Die Radiolarien, lS62.—Jdem, Report on the Radiolaria, in Scient. Results Cliallenger Exped., Zool., vol. xviii., 1887. — Hertwig, R., Der 42 PROTOZOA— RHIZOPODA phylum i more nuclei, vacuoles, alveoles, granules, oil- globales and sometimes crystals and surrounded by a capsule-membrane perforated by pores or pylae ; and (2) an outer jelly-like extra-capsulum, the sarcode of which emits pseudopodia. The individuals lead usually an isolated existence, and are only rarely united in colonies. Most Radiolarians secrete skeletons composed of either bars or spicules of acanthine (an organic substance allied to hörn or chitin) or silica, or they build an exceedingly delicate lattice-work composed of transparent amorphous silica. Only the latter forms are known in a fossil State, and owing to their minute size, are commonly indiscernible except with the aid of the microscope. Haeckel divides the Radiolaria into four suborders, as follows : — A. Acantharia. — Capsule-membrane uniformly perforated ; skeleton com- posed of acanthinic spicules. Unknown in fossil State. B. Spumellaria. — Capsule-membrane single, pores distributed all over; skeleton siliceous, spherical or discoidal, sometimes wanting (Fig. 43). Fig. 41. Silurian and Devonian Radiolarians : A, Cenosphaera macropora Rüst. Ordovician ; Cabrieres, Langnedoc. B, Staurolo)iche micropora, Rüst. Ordovician ; Cabrieres. C, Caryosphaera groddecJd Rüst. Upper Devonian ; Schäbenholz, near Elbingerode, Harz Mountains. D, Lithocampe tschernytscJie%ni Rüst. Devonian ; Ural. Magnifted 100 to 120 diameters (after Rüst). C. Nasselaria. — Capsule-membrane single, perforated only about the otäl pole ; skeleton siliceous, helmet- or cap-shaped, conformation of poles dissimilar (Figs. 44, 45). D. Phaeodaria. — Capsule-membrane double, perforated by one main opening prolonged into a tubulus, and by a few smaller accessory openings. A dark pigment body (phaeodium) constantly present in extra-capsular sarcode. Skeleton commonly consisting of hollow siliceous spicules disposed in flask-shaped or variously shaped frameworks. Unknown in fossil state. Radiolarians are exclusively marine organisms, and are found at all bathy- metric zones. They occur in vast numbers, especially in tropical seas, swimming on the surface, as well as at medium and even abysmal depths. Particularly between 2000 and 4000 fathoms in depth, extensive deposits of " Radiolarian mud " have been found, the composition of which is largely silica with a small percentage of carbonate of lime. Organismus der Radiolarien. Jenaische Denkschr., 1879, vol. ii. — Stöhr, E., Die Eadiolarien- Fauna von Grotte in Sicilien. Palaeoutogr., 1880, vol. xxvi, — Rüst, D., Radiolarien aus Gesteinen des Jura. Palaeontogr., 1885, vol. xxxi. — Idem, op. cit., vols. xxxiv. xxxviii. and Iv. — Dreyer, F., Die Tripoli von Caltanisetta. Jenaisclie Zeitsclir. f. Naturw., 1890, vol. xxiv. — Cayeux, Z., Les Preuves de l'existence d'organismes dans le Precambrien. Bull. Soc. Geol. France, 1894, vol. xxii. — Vinassa de Regny, P. E., Radiolarie delle flaniti titoniane di Carpena (Spezia). Palaeont. Italica, 1899, vol. iv. — Hinde, G. J., Radiolaria in Devonian Rocks of New South Wales. Quart. Journ. Geol. Soc, London, 1899, vol. \v.—Ideni, Radiolaria from the Triassic of the Dutch Fast India Archipelago. Jaarb. Mijnwezen Nederl. Ooost India, 1908, vol. xxxvii. — Squinabol, S., Radiolarie cretacee degli Euganee. Padova, 1904. — Principi, P., Contributo allo studio dei radiolari miocenici italiani. Boll. Soc. Geol. Ital., 1910, vol. xxviii. ORDER TU RADIOLARIA 43 The diversity of form exhibited by Radiolarians is very remarkable, and Fig. 42. Carbotiiferous, Jurassic, and Cretaceous Radiolarians: A, Staur(u:ontium ina^qualeRnsit. Carboniferous ; C, Xiphodictyä ac\äa Rüst. In coprolite Sicily.' B, Trochodisctis nicholsoni Rüst. Carboniferous; Harz. -, -^ . «„^-r^nA. from Lias ; Ilsede, Hanover. D, Hymenütatnm rotvnditm Rüst. In coprolite from Cretaceous ; /Uli, öaxony. the Identification of their microscopic siliceous skeletons is impossible without the aid of special literature. Contrary to formerly current ideas, the geological antiquity of the Radiolarians is very great ; and they also play an important part in the cora- position of many siliceous and cal- careous-siliceous rocks (quartzites, hornstone, Jasper, phyllites, Aptychen- schiefer, etc.). Ac- cording to Barrois they are the oldest known animal or- ganisms, since the Spumellaria (Mono- sphaeroidae) occur plentifully in the bituminous quartz- ites of Brittany, interbedded with pre - Cambrian gneiss. Although the group is still very imperfectly known, yet, according to Rüst, fossil Radio- laria are by no means less abundant and less diversified than the Recent Recent and Tertiary Spumellarians : A, Adinomma astenunnthium Haeck. Recent ; Messina. B, StylodictyamulHspina J1aec\i. Recent ; Messina. C,Helio- discus Inimboldti Ehrbg. Barbados eartli (Miocene) ; Barbados. D, Haliomma dixlphos Ehrbg. Miocene marl ; Caltan isetta, Sicily. E, Astromma aristotdis Ehrbg. Miooene ; Barbados. Only in exceptional 44 PROTOZOA— EHIZOPODA PHYLUM I Fig. 44. Recent and Tertiary Nasselarians : A, Podocyrtis schcmiburgki Ehrbg. Tertiary marl ; Barbados. B, Cyrtocalpis amphora Haeck. Recent ; Messina. C, Bothryocampe hexathalamia Haeck. Recent ; Mediterranean. D, Fetalospyris foveolata Ehrbg. Tertiary marl ; Barbados. instances (Miocene of Barbados, Oran, Sicily) have the skeletons been preserved unaltered, and still consist of amorphous silica. In the older rocks the silica has usually become dissipated in the matrix, being replaced by lime carbonate, iron, or some colouring agent ; in other cases the quartz has become crypto- crystalline, or replaced by a calcite pseudomorph. The Cambrian Griffel- schiefer of Sonneberg in Thuringia contain poorly preserved Sphaeroidea ; the usually dark, though some- times red or light-coloured Ordovician strata of Lan- genstriegis in Saxony, and of Rehau and Stehen in Franconia, the red Jasper of Abington, Scotland, and the Ordovician sili- ceous rocks of Cabrieres in Languedoc, are more or less rieh in Radiolarian remains belonging exclusively to the Spumellaria (Fig. 4:1, A, B). From the Devonian Jasper of Siberia, the siliceous schists of Hesse and Nassau, and the manganiferous quartzite of Elbingerode in the Harz, and other places, Rüst has described forty-six Spumellarian species and seventeen Nasselarian (Cyrtoidea). The Lower Carboniferous quartzites, phyllites, adinole and Jaspers from the Harz (Culm formation), Ural district and Sicily have yielded 155 species, of which thirty-six belong to the Nasselaria. In general the Paleozoic Radiolarians are remarkable for their relatively large size and excellent preservation. The Triassic appears to be destitute of Radiolarians except in the Alps, where they are abun- dant in the hornstone and siliceous lime- stone of the Buchen- stein beds of Hungary, and occur less fre- quently in the Rei- fling limestones, in the Wengen beds of Storzic in Carniola, in the marls of St. Cassian, and in the siliceous limestone of the Rötheistein, near Aussee, etc. They are usually associated here with the remains of Sponges and Foraminifera. In the silicified coprolites of the Lias, found at Ilsede, Hanover, Radiolarians are very common; they are somewhat less frequent in the limestones of the Lower Lias on the Schafberg in Upper Austria. Certain hornstone beds of Middle Jurassic age, found at Piszke, Hungary, the Upper Jurassic pudding-stones of Fig. 45. Tertiary Nasselarians from Barbados : Ä, AntJwcyrtis mespihis Ehrbg. B, Lychnocanium lucerna Ehrbg. C, Dictyomitra montgolfieri Ehrbg. D, Eucyr- tidium elegans Ehrbg. E, Pterocodon campana Ehrbg. ORDER III RADIOLARIA 45 Cittiglio, near Laveno on Lago Maggiore, and numerous Tithoniaii jaspars, as well as the Alpine Aptychus beds, are charged with Eadiolarians ; here the Nasselaria are nearly as plentiful as the Spumellaria. The Lower Cretaceous (Neocomian) of Gardenazza has yielded but few forms. On the other hand, coprolites from the Gault, found near Zilli in Saxony, and Lower Cretaceous clay marls in Manitoba, Canada, as well as from Upper Cretaceous marls of Haldem in Westphalia, and Vordorf in Brunswick, contain excellently preserved skeletons in greater or less abundance. Even the flinty concretions of the Upper Chalk sometimes contain them, although in a poor State of preservation. Certain Eocene hornstones in Italy, according to Pantanelli, are filled with Radiolarian remains, while in the Flysch they are also very profuse in some localities, although usually poorly preserved. By far the most noted occurrence of fossil Radiolarians is in the siliceous " Barbados earth," of Miocene age, in which Foraminifera are also very con- spicuous ; while the " tripoli " of Grotte, Caltanisetta and Girgenti in Sicily, of Oran, Aegina, Zante, the Nikobar Islands and other localities (Miocene and Pliocene), is scarcely less noteworthy. Ehrenberg has described 278 species from Barbados alone, and from Sicily Stöhr has described 118 species, most of which belong to still extant Spumellarian, Nasselarian and Phaeodarian genera. Phylumll. COELENTERATA. CoELENTERATES or Zoophytes are free-swimming or attached aquatic animals of very variable form and size. They difFer from the Protozoa in having multicellular bodies with distinct organs ; and from all higher classes in the absence of a definite body-cavity. In the subphylum Porifera there is a simple or usually complex System of digestive sacs, with inhalent pores in the body wall and one or many exhalent pores or oscula, and no stinging cells or tentacles. The two other subphyla, Cnidaria and Ctenophora, exhibit a more or less pronounced radial symmetry, have no inhalent pores and no special exhalent opening in the body wall, but a large mouth opening conducts into a gastrovascular cavity. Food is taken in and the excreta and sexual elements are voided through the mouth opening. Stinging cells and usually tentacles are present in the two last-named divisions. The body consists of two layers of cells — an ectoderm and entoderm — and usually also a third layer, the mesoderm. The ectoderm in the Cnidaria often secretes a calcareous or horny skeleton, but in the Porifera the horny, siliceous br calcareous skeletal elements are the product of the mesoderm. Eeproduction is either sexual or asexual, or, in the Hydrozoa, an alterna- tion of generations may occur. The process of budding or self-division gives rise to polyzooid colonies, in which the zooids subsist in intimate relationships with one another, and sometimes Institute a physiological division of labour. R. Leuckart was the first to recognise the Coelenterates as constituting a distinct structural type of animals and separated them from the Echinoderms, with which the older systematists had associated them under the general term of Radiates or Actinozoa. The Coelenterates are divided into three principal groups or subphyla : Porifera, Cnidaria and Ctenophora. Of these only the first two have skeletons and have left traces in the rocks. Subphylum L Porifera Hogg. The Porifera or Sponges are sessile, aquatic animals of extremely variable form. The body consists of a single layer of pavement-cells forming the ectoderm, a single layer of collared epithelial cells constituting the entoderm, and usually a strongly developed mesoderm, which latter comprises the bulk of the soft parts (including all the organs, muscles, sexual elements and nerves), and almost invariably secretes a hard skeleton. The latter may consist of horny sponge-fibres, or of regularly disposed siliceous or calcareous skeletal elements. The whole body is ramified by a canal-system, and the outer epithelial layer is perforated by countless minute, dermal pores for the entrance of water laden with food-particles. The pores communicate by means of fine 46 CLASS I SPONGIAE 47 f Hei curreiit canals with subdermal ciliated Chambers, from which larger exciirrent nals coiiduct the water and food or excreta through the body, and generally en into a wide, exhalent opening called the cloaca or paragaster. Stinging 11s, tentacles and radial mesenteries are absent. The Porifera comprise but one class, the Sponges. Olass 1. SPONGIAE. Sponges.i Sponges are remarkable for their extreme variability in external form and size ; they lead either an isolated existence, or are united in colonies of cylindrical, tubulate, pyriform, fungns-like, bulbous, spherical, compressed, foliate, umbel-, bowl- or beaker-shaped, or of botryoidal form. They are long- or short-stemraed, or a peduncle may be absent ; sometimes the stock is branching, and the arms may be either separate or interlaced so as to form networks. Nothing is less stable than the outer conformation, which varies excessively aecording to the Situation and other physical conditions, and whose systematic importance, accordingly, is very slight. The size is also extremely variable, ranging from that of a pin-head to 1 J metres in diameter. Sponges are invariably sessile in habit, being attached either by means of a stem or a bündle of anchoring spicules, or they may be simply encrusting at the base. The canal-system by which the whole body is traversed, is extremely com- plicated in thick-walled, but simple in thin-walled sponges. A distinction is recognised between incurrent or inhalent, and excurrent or exhalent canals. In the terminology proposed by Rauff, inhalent canals are designated as epirrhysa, and exhalent canals as aporrhysa ; the former terminate on the periphery in ostia (not to be confounded with the finer dermal pores), while the latter terminate on the cloacal surface in postica (again not to be con- founded with gastral pores). Postica are usually larger than ostia, and differ from them in form and arrangement. ^ Literature : ^. On receut Sponges : — Schmidt O., Die Spoiigien des Adriatischen Meeres. Leipzic, 1864-66. — Ide7)i, Die Spongien der Küste von Algier. Leipzic, 1868. — Idem, Die Spongien des Meerbusens von Mexico. Jena, 1879-80. — Haeckel, £., Die Kalkschwämme, 1872. — Schulze, Fr. Iß., Untersuchungen über den Bau und die Entwicklung der Spongien. Zeitschr. f. wiss. Zool., 1876-80, vols. xxvii. - xxx. — Report on the Hexactinellida. Scient. Results Challenger Exped., Zool,, vol. xxi., 1887. — Vosmaer, G. C. J., Spongien {Porifera), in Bronn's Classen und Ordnungen des Tierreichs, 2nd ed., 1882-87, vol. iii. — Lendenfeld, R., A Monograph of the Horny Sponges. London, 1889. B. On fossil Sponges : — Ooldfuss, A., Petrefacta Germaniae, vol. i., 1826-33. — Michelin, H., Iconographie zoophyto- logique, 1840-47. — Fromentel, E. de, Introduction ä l'etude des eponges fossiles. Mem. Soc. Linn. Normandie, 1859, vol. xi. — Roemer, F. A., Die Spongitarien des norddeutschen Kreidegebirges. Palaeontographica, 1864, vol. xiii. — Zittel, K. A., Ueber Coeloptychium. Abhandl. k. bayer. Akad., 1876, vol. xiii. — Studien über fossilen Spongien, i., ii., iii., ibid., 1877, vol. xiii. (translated by Dallas in Annais and Mag, of Nat. Hist. for 1877, 1878, 1879). — Beiträge zur Systematik der fossilen Spongien, i,, ii,, iii., Neues Jahrb. für Mineral. 1877, 1878, l^l^.—Quenstedt F. A., Petrefactenkunde Deutschlands, 1877, vol. v. — Sollas, W. J., Quart. Journ. Geol. Soc. 1877-80, vols. xxxiii. - xxxvi, — Hinde, G. J., Catalogue of fossil Sponges of British Museum, London, 1883.— Monograph of British fossil sponges ; Palaeontographical Society, 1887, 1888, 1893. — Rmtff, H., Palaeospongiologie ; Palaeontographica, 1893-94, vols. xl., xli. (contains füll bibliography). — Schrammen, A., Beitrag zur Kenntniss der obersenonen Tetractinelliden. Mittheil. Roemer. Museum Hildesheim, 1899-1903, Nos. 10, 14, 15, 19.— Hall, J. and Clarke, J. M., A Memoir on the Palaeozoic reticulate Sponges constituting the family Dictyospongidae. N. Y. State Mus. Mem. ii., 1898. Earlier contributions by same authors in loth and 16th Reports N. Y. State Geologist, 1895-96. — Schrammen, A., Kieselspongien der oberen Kreide von Nord Westdeutsch- land. Palaeontogr. 1910, Supplem. vol. v. — Kolb, R., Kieselspongien des schwäbischen weissen Jura. Op. dt., 1911, vol. Ivii. 48 COELENTERATA— PORIFERA phylum ii The water enters through the dermal pores, and passes through the incurrent canals into ciliated Chambers, which are lined with epithelial cells. From these it is conveyed through all parts of the body by means of the frequently branching excurrent canals, which open into a sac-like, tube-like or funnel-shaped cloaca. The exhalent opening of the latter is termed the osculum. Extremely thin-walled sponges have no cloaca, osculum or branch- ing canal-system, but the excurrent canals terminate directly in small openings situated on the upper surface of the body. The cloaca when present is often of considerable depth, although sometimes shallow, or reduced to a mere sac-like Prolongation of the osculum. Forms with a large and deep cloaca are regarded as Single individuals, those with numerous cloacae and oscula as colonies. But since all the cloacae of a colony communicate by means of canals, while the oscula are never surrounded by a crown of tentacles, it is often difficult to distinguish between large excurrent canals and true cloaca, and hence also between individuals and colonies. Reproduction is either sexual or asexual. In the first process the fertilised ova complete a tolerably regulär segmentation, develop into a gastrula, pass out through the osculum, and attach themselves to some foreign object. Asexual reproduction takes place by budding, the young buds remaining attached to the parent individual, and thus giving rise to colonies. Re- production by means of fission forming new colonies is of rare occurrence. The great majority of sponges secrete a skeleton composed either of horny fibres or of siliceous or calcareous spicules, or they incorporate foreign bodies into their framework. Only a few Recent forms (Myxospongiae) are without a skeleton. In the horny sponges (Ceratosponyiae) the skeleton consists of anastomosing and reticulated fibres of spongin, an organic nitrogen Compound resembling silk. The fibres are either solid, or they contain an axial canal, which is sometimes cored with foreign bodies, such as sand-grains, fragments of sponge-spicules, foraminifers, radiolarians, etc. Siliceous spicules are sometimes encased in horny fibres, sometimes occur detached in the cellular tissues, or are interlaced and Consolidated with one another in various ways to form a supporting framework. In each genus the skeleton is composed of but a single type, or at the most of but a few regularly repeated varieties of siliceous bodies, which are called the skeletal elements. In addition to these there occur more or less abundantly, especially on the outer surface and in the cloacal and canal walls, extremely delicate flesh-spicules, usually of small size and of great diversity of form. The flesh-spicules are as a rule destroyed during fossilisation. All the siliceous skeletal elements are secreted by nucleated cells, and are composed of concentric layers of colloidal silica, deposited usually about a slender axial canal. In some spicules, notably those having spherical or stellate contours, the axial canal is wanting. It is very delicate in fresh spicules, but becomes enlarged by maceration, and in fossil speciraens it is often coarsely calibrated. The multitudinous varieties of siliceous skeletal elements (Fig. 46) may be grouped into a few fundamental types, as follows : — {a) Uniaxial spicules or monaxons (Fig. 46 ^'^^) and (^'^"^^). Straight or bent, smooth, prickly or knotty, bevelled, sharpened or truncated needles, rods, hooks, clasps, pins and anchors (amphidiscs). They invariably contain an axial canal, which may be either entirely sealed up, or open at one or at both ends. }S I SPONGIAE 49 I ^^m (b) Tetraxial spicules or tetraxons (Fig. 46 ^^). The normal form is ^Waracterised by four equal rays intersecting like the bisectrices of the plane angles of a regulär tetrahedron. Triaxial forms result from the occasional abortion of one of the rays. One of the rays may become elongated or other- wise modified so as to form anchors (triaens) with three simple or furcate hooks (Fig. 46 18-23^^ Three of the rays may be numerously divided or foliately expanded so as to produce forms resembling thumb-tacks {trichotriaens, phyllo- triaens) ; atrophy of the fourth ray in the last-named form reduces the spicule to a delicate siliceous disk (Fig. 46 ^). A peculiar forking of the shaft gives rise to candelabras or amphitriaens, while other modifications may produce umbellate spicules (Fig. 46 ^^), etc. Certain skeletal Clements of the Lithistids (Figs. 48-63) may be regarded as irregulär tetraxons {desmoms), in which the extremities of the four rays are Fio. 46. Varlous forms of Sponge spicules from the Upper Cretaceous of Haldem, Westphalia ; magnified 25 diameters. 1-6, Uniaxial rod.s and needles. 7-9, Uniaxial siliceous elements with coarse axial canals. 10-13, Uniaxial cylinders and spheres. 14, Microspined spicule. 15, Clasp-hook flesh-spicule. 16, Bispatulate tlesh-spicule. 17, llegular four-rayed spicule (chevauxde frise). 18-21, Trifld anclior-shaped spicules. 22-23, Anchors with furcate head-rays. 24-25, Irregulär four-rayed skeletal elements. 26, Umbel-shaped spicule. 27, Six-rayed spicule. 28, Polyaxial siliceous disk. prolonged in knotty, root-like excrescences, or in which, owing to the un- symmetrical growth, branching or atrophy of one or more of the arms, extremely irregulär forms are produced ; for these a special terminology has been devised by Rauff. (c) Hexactinellid spicules (hexadins or triaxons) (Figs. 65-70). The ground-form is an axial cross with six equal arms intersecting at right angles like the axes of a regulär octahedron. Atrophy of one or more of the rays may result in pentaxial, tetraxial, triaxial, or even clavate forms, without their real character becoming entirely obliterated. Bifurcation or other modifica- tions of a number or of all the rays produce beautifully formed siliceous structures highly characteristic of the group IlexadineUida, which resemble candelabras, double-headed anchors, fir-trees, pitch-forks, rosettes, etc. The fusion of juxtaposed hexactins produces more or less symmetrical latticeworks with cubical interstices. VOL. I K 50 COELENTERATA— POEIFERA phylum i {d) Anaxial or polyaxial bodies of spherical, cylindrical, stellate or discoidal shape, which are not derivable from either of the tliree ground-forms, occur in only a few varieties of recent and fossil siliceous sponges. Calcareous skeletal elements are much less complicated, and are generally smaller and more perishable than the siliceous. Their form is either triaxial (triods), tetraxial (tetraxons), or nail-shaped (monaxons). The triaxial and tetraxial spicules are very rarely forked or otherwise modified. Each skeletal element behaves optically like a single calcite crystal ; axial canals are absent. The skeletal elements in sponges are arranged chiefly with reference to the circulation of water through the canal-systems. In thin-walled forms they are more or less closely crowded together, and are often regularly oriented in the soft parts ; in other forms they are encased in horny fibres, or are packed in between the canals ; in still others they are united to form an irregulär frame- work, or may be welded together in a regularly reticulated scaffolding. The horny fibres are totally destroyed during fossilisation ; calcareous spicules are often wholly or partially dissolved, or are replaced by infiltrating lime carbonate, and assume a dense fibrous appearance (Fharetrones). Like- wise in siliceous sponges the skeletal elements are rarely preserved unaltered ; as a rule the originally colloidal silica becomes crystalline, or is dissolved and removed. The cavities thus formed may subsequently become filled with infiltrating quartz, limonite or most commonly with lime carbonate. In this manner the skeletons of fossil siliceous sponges are converted into calcite, and, contrariwise, spicules that were originally calcareous may become silicified. Hence the distinction between siliceous and calcareous sponges in the fossil State depends entirely upon morphological characters, and not at all upon the chemical composition of the preserved parts. Sponges are divided into four subclasses : Myxospongme, Ceratospongiae, Silicispongiae and Calcispongiae. The latter group Stands in sharp contrast to the other three, which are connected by intermediate forms, and constitute together a 'group of equal value with the calcareous sponges. Skeletal elements are absent in the Myxospongiae, whose bodies are composed entirely of soft cellular tissues. The Ceratospongiae also lack imperishable hard parts, the spongin fibres being entirely destroyed during fossilisation. The reputed horny sponges from the Trias (EhizocoraUum), Jura and Cretaceous (Spongites, Saxonicus, Paramudra, etc.) are either of inorganic nature or are zoologically indeterminate. All fossil sponges, therefore, belong either to the Silicispongiae or the Calcispongiae. The oldest forms are found in the Cambrian ; in the Trias, Jura and Cretaceous they are very abundant. Subclass 3. SILICISPONGIAE. Siliceous sponges. Skeleton composed either exclusively of siliceous elements, or of horny fihres enclos- ing siliceous spicules. Order 1. MONAGTINBLLIDA Zittel. {Monaxonia F. E. Schulze.) All skeletal elements uniaxial. The Monactinellida include the majority of existing marine sponges, most of which occur at moderate de23ths ; and also the few fresh-water forms ORDER II SILICISPONGIAE— TETRACTINELLIDA 51 (Spongilla) that are known. The skeleton, as a rule, is composed like that of the horny sponges, of anastomosing spongin fibres, which either encase rod-like spicules, or coiitain quantities of iiniaxial siliceous elements ; some- times the latter are also present in the soft parts. In each genus there are usually either one or but few varieties of siliceous elements present, which are uni- formly distributed throughout the body. Needles, hooks, crotchets, cylinders, spindles, amphidiscs and the like occur in great diversity. Owing to the decomposition of the horny fibres during fossilisation, and the fact that the skeletal elements are never fused together, the latter become detached and scattered in all directions. While Monactinellid spicules are very common in certain formations, they are rarely united in the form of coherent skeletons, and are only capable of generic determination when their form is sufficiently characteristic, as in Eenieria, Esperia, etc. The basal beds of the Alpine Lias often contain considerable hornstone, which is sometimes com- pletely filled with rod-shaped spicules. In various Cretaceous and Tertiary horizons Monactinellid spicules are also enormously abundant. Hinde has described a Climacospongia from the Silurian of Tennessee, in which the skeleton consists of spicules arranged in longitudinal rows, and connected by transversely disposed elements. The spicules were probably originally enclosed in horny fibres. The Clionidae secrete pin-shaped siliceous elements which are also encased in horny fibres, and Recent sponges of this family bore labyrinthic passages in the shells of mollusks. Fossil sponge-borings are also common. Detached spicules of Eenieria, Axinella and Haplistion have been described by Hinde from the English Carboniferous Limestone. Order 2. TETRACTINELLIDA Marshall. (Tetraxonia F. E. Schulze.) Skeleton composed of regulär tetraxons which are generally combined with uni- axial, polyaxial or heteraxial siliceous bodies. The skeletal elements occur detached throughout the soft parts, and are never united to form a connected frame- work. The most common forms of skeletal elements are normal tetraxons, anchors with simple or furcate prongs, spheres and stellate bodies. In certain genera (Geodia) the large anchors and cylinders are disposed in radiately arranged fasci- cles, and are surrounded by a thick layer of anaxial spheres. Detached Tetractinellid spicules associated with Monactinellids occur more or less abundantly in the Carboniferous Limestone, the Alpine Infra- Lias, the English Neocomain, the Deister Sand- stone (Hils), the Upper Cretaceous of Haldem and Coesfeld in Westphalia, and in the Tertiary and Pleistocene formations. The skeletal elements are preserved in their natural position in the genera Ophiraphidites Carter ; Tethyopsis Zittel (Fig. 47), Pachastrella Schmidt, Stolleya and Cephaloraphidites Schrammen. Fio. 47. Tethyopsis steinmanni Zittel. Upper Cretaceous ; Ahlten, Hanover. 14/^. 52 COELENTERATA— PORIFEEA PHYLÜM II Orders. LITHISTIDA Schmidt. Massive, thick-walled, siliceous sponges, usually with complicated canal-system. Skeleton composed of irregulär tetraxons or monaxons {desmoms) which develop nodose or root-like hranches either at the extremities or all along the shaft, and are firmly united hy zygosis. Symmetrical, tetraxial, uniaxial or polyaxial dermal and flesh-spicules also present. The Lithistids are closely related to the Tetractinellids, and in the opinion of many zoologists, should be embraced in the same order with them. The Lithistids are peculiarly well adapted for preservation, owing to the massive stony character of their skeletons ; and their remains occasionally form thick deposits, especially in the Jurassic and Cretaceous. Their outer configuration is extremely variable ; most commonly it is crateriform, cup-shaped, pyriform, globular, bulbous or plate-like; while the body is attached either by the base or by means of a peduncle. The canal-system varies greatly in different genera, but is usually well developed and more or less complicated. The four-rayed skeletal elements are interlocked by means of the root-like branching ends of the rays, and the points of intersection (nodes) with the ends of adjacent uniaxial spicules are thickened into balls. The usually irregulär uniaxial skeletal elements are interlaced on all sides by means of root-like processes. Dermal and flesh-spicules are preserved only under exceptionally favourable conditions, but are invariably present in recent genera, and furnish valuable differential characters. The Classification of fossil Lithistids is based wholly upon the skeletal elements and canal- systems. Five principal groups are recognised, whose subdivision into families need not concern us at present : — Tetracladina, Euiaxicladina, Anomo- cladina, Megamorina, and Bhizomorina. Existing Lithistids occur most abun- dantly at depths ranging between 100 and 400 metres, but are occasionally found as deep as 1800 metres. Suborder 1. TETRACLADINA Zittel. Skeletal elements composed of four usually equal rays, each of which endoses an axial canal, and has extremities ferminating in root-like Strands or processes ; the spicules are intertwined to form an open mesh- work. Dermal spicules either grapnel-like te- traxons, frequently with furcate prongs, or dis- coidal with entire or lohate margin; or they are nail - shaped or cylindrical monaxons. Fio. 48, Aulocopiurfi aurantium Oswald. Diluvium ; Sadowitz, Silesia. a, Example in I/o natural size ; h, Skeleton magnified 60 diaraeters. The skeletal ele- ments of the Tetra- cladina are usually symmetrical tetrax- ons, whose four smooth, more rarely tuberculate or knotty rays intersect ap- proximately at an angle of 109^°. They occur in the Cambrian and Silurian, »IIDER TU SILICISPONGIAE— LITHISTIDA 53 are very scarce in the Upper Jurassic (Protetraclis), but abundant in the Cretaceous, Tertiary and Eecent periods. A'ulocopium Oswald (Fig. 48). Hemispherical or bowl-shaped with short peduncle ; inferior surface covered with a dense, wrinkled, siliceous skin. Fig. 49. Callopegma acaule Zitt. Senonian ; Ahlten, Hanover ; a, Specimen in 3/^ natural size ; h, Skeleton magnified 40/j ; c, Portion of periphery, 2/j ; d, Same magnified 40/jj and showing anchors with furcate head-rays. Cloaca central ; sponge body with numerous arched canals parallel to contour of periphery, and with finer radial canals leading from exterior to cloaca. Skeleton composed of irregulär smooth-rayed tetraclons with root-like branch- ing extremities, disposed in rows parallel to the radial canals. Occurs (usually replaced by calcite) in the Ordovician of the Kussian Baltic Sea Provinces, Fig. 50. Phymatella tuberosa (Quenstedt). Quadratenkreide f'Upper Senonian); Linden, near Hanover. a, Sponge, 1/2 natural size ; b, Outer surface, 1/1 ; c, Skeletal element, •'*Vl ; d, Spicules from stalk portion, 50/j. Fig. 51. Siphonia tulipa Zitt. Greensand ; Blackdown. A, Longitudinalsection, natural size. B, Sponge with ])eduncle and root, 1/2 natural, size (after Sowerby). Ordovician of Illinois, and Silurian of Gotland. Also in erratic blocks on the plains of Northern Germany, usually chalcedonised. Archaeoscyphia Hinde. Cambrian. Callopegma Zittel (Fig. 49). Bowl- or funnel-shaped, short-stemmed, thick- walled. External surface perf orated by smaller, internal by larger canal-openings 54 COELENTEEATA— PORIFERA PHYLÜM II (ostia and postica). Skeleton composed of smooth-rayed tetraclons, the digitate extremities of which are inflated into balls. Dermal spicules in the form of anchors and rods. Upper Cretaceous. Phymatdla Zittel (Fig. 50). Upper Cretaceous. Polymaraphinina, SoUasella, Pseudoplocoscyphia and Craterella Schrammen. Upper Cretaceous. Fig. 53. Skeletal element of Jerea quenstedti Zittel, showing brancliing extremities of rays. Qnadratenkreide; Lin- den, near Hanover. ■^^/i. Fig. 54, PUnthosella sqiiamosa Zittel. Skeleton magnilied 80 diameters. Senonian ; Alilten, Hanover. Fig. 52, Jerea pyriformis Lamx, Greensand ; Kelheim, Ba- varia, 1/2 natural size. Siphonia Park. (Fig. 51). Fig-, pear- or apple-shaped, with a long or short ped- uncle. Body with deep cloaca, into which arched canals running parallel with the periphery, to- gether with numerous fine radial canals, con- duct. Skeleton com- posed of smooth-rayed, branching dichotriders. of monaxons and grapnels, Abundant in Middle and Upper Cretaceous, Hallirhoa Lamx. Like the preceding, but invariably short -stemmed. Body pyriform and lobate, owing to a number of deep constrictions. Cenomanian. Jerea Lamx. (Figs, 52, 53). Body pyriform, flask-shaped or cylindrical, with truncate or depressed summit, in which a number of tube-like canals, vertical in the central portion but arched in the peripheral, terminate. Cross- ing the latter are finer radial canals, Skeleton composed of tetraclons and dichotriders. Common in Middle and Upper Cretaceous. Rhagadinia rimosa Roemer, natural size ; b, Skeleton, 40/j Spicule of dermal layer, 4o/j, c, Lobate disk from dermal layer, 40/^ ; Dermal spicules in the form SILICISPONGIAE— LITHISTIDA 55 Folyjerea From. ; AstroclcuUa, lliecosiphonia, Colymmatina Zitt. ; Turonia Mich. ; riinthosella Zitt. (Fig. 54). Cretaceous. Discodermia Boc. ; Rhacodiscula Zitt., etc. Cretaceous and Tertiary. Rhagadinia Zittel (Fig. 55). Auricular, plate- or bowl-shaped, short- stemmed. Both surfaces traversed by irregulär branching furrows, in which the canalicular Ostia are situated. Skeletal elements four-rayed, sometimes uniformly or only distally covered with tuberculous knobs, and with digitate extremities. Dermal spicules in the form of six-lobed disks, provided with a short shaft, and minute, multifid tetraclons. Upper Cretaceous. Suborder2. EUTAXICLADINA Rauff. Skeleton composed of four-rayed spimles with three equally developed simple or hifurcate rays which terminale distally in root-like fibres ; and one abbreviate, inßated fourth ray (ennomoclon). Axial canals probably in all of the rays. Skeletal elements invariably arranged in either parallel or alternating rows, and unifed by zygosis into a network with triangulär or irregulär meshes ; spicular nodes greatly inflated. Nearly all the genera are Silurian ; a few (Mastosia, Lecanella) occur in the Upper Jurassic. ^yCS^^ Astylospongia praemorsa (Goldf.). In erratic block from Mecklenburg, a, Sponge, partially cut into, natural size; b, Skeleton, 12/^ ; c, Portion of same highly magnified. Astylospongia Koem* (Figs. 56, b7a). Spherical, with shallow depression on the summit ; base evenly rounded, unattached ; probably fastened by means of anchoring fibres. Large-sized canals directed parallel to periphery in the outer portion of the body, vertical in central portions ; besides these there are numerous fine radial canals which terminate in pores all over the periphery. Skeletal elements with four smooth elongated rays, one or all of which brauch dichotomously just above the junction with the shorter arm. Spicular nodes thickened into large knots. Ordovician of the Russian Baltic Sea Provinces, and Silurian of Sweden and North America (not- ably in Tennessee), usually chalcedonised. Also in erratics in the Diluvium of Northern Germany. Fig. 57. a, Detached skeletal dement of Astylo- fponriia, i'-^o/j ; b, Detached skeletal ele- ment of Hindia, 80/^ (after Kauff). 56 COELENTERATA— POEIFEKA PHYLUM II Caryospongia, Carpospongia RaufF. Ordovician and Silurian ; Europe. Falaeomanon Roem. {Astylomanon RaufF). Like Astylospongia, but bowl- shaped, with shallower and wider cloacal depression. Entire . surf ace covered with pores. Silurian ; North America. P. cratera Roem. Caryomanon, Carpomanon RaufF. Silurian ; North America. Hindia Duncan (Fig. 57&). Body spherical, with perforate periphery, traces of attachment wanting. All canals radiate from the centre outward. Skeletal elements composed of three simple rays beset with prickly tubercles, and a reduced button-like fourth arm. All spicules regularly disposed in rows parallel with the radial canals. Silurian ; North America. Neohindia Schrammen. Upper Cretaceous ; Germany. Suborder 3. ANOMOCLADINA Zittel. {Didymmorina RaufF.) Skeletal elements composed of short, smooth rays with spherically inflated ends which give off three, four or more simple or digitale hranches ; the latter are united hy zygosis with processes of adjacent rays ; axial canals simple. Dermal spicules rod-shaped monaxons. Upper Silurian to Recent. Anomoclonella, Pycnopegma RaufF. Silurian ; North America. Cylindrophyma Zittel (Fig. 58). Body cylindrical, thick-walled, attached ; cloaca wide and tube-like, receiving numerous radial canals, and extending Cylindrophyma milleporata (Goldfuss). Upper White Jura ; Hoclisträss. A, Two specimens, 1/2 natural size. B, Skeleton magnified 30 diameters. C, Detached siteletal element of Cylindrophyma, 60/^ (after Rauflf). down as far as the base. External surface perforated by fine ostia. Common in Upper Jurassic. Melonella Zittel. Skeleton apple-shaped or hemispherical, with broad base, or provided with very short .peduncle ; base covered with wrinkled siliceous skin. Cloaca deep, funnel-shaped. Coarser canals arched, parallel with peri- phery ; finer incurrent canals radially directed. Upper Jurassic. M. radiata (Quenstedt). ORDER III SILICISPONGIAE— LITHISTIDA 57 Suborder 4. MEGAMORINA Zittel. (Rhabdomorina RaufF.) Usually large-sized, elongated, smooth, bmt, loosely interlocking, irregularhj hranching, or only terminally forked skeletal Clements with simple axial canals ; interspersed among which small, radiciform, numerously hranching Clements {rhizomorins) are occasionally present. Dermal spicules uniaxial or grapnel-shapcd. Ordovician, Silurian, Carboniferous, Jurassic, Cretaceous and Recent. Saccospongia Rauff. Silurian. Megalithista Zittel. Upper Jurassic ; Nattheim. Dory derma Zittel (Fig. 59). Sponge-body cylindrical, simple or branching, pyriform or compressed, with a number of larger canals running parallel with the body axis, and numer- ous smaller radial canals. Skeletal Clements large, bent and divided into two or more simple branches. Dermal spicules in the form of three - fluked an- chors. Upper Cretaceous ; Nor- thern Germany, England and Fig. 59. France AcCOrd- Doryderma dichotoma (Roemer). Upper Cretaceous. a, Sponge, natural size ; b, TT- J Dermal layer, 2/^; c, Bündle of skeletal elements, lo/j ; d, Skeletal element and ing to xlinde, also several dermal spicules with furcate, anchor-shaped head-rays, 30/^. Carboniferous. Gartcrclla Zittel ; Astcroderma Schrammen. Cretaceous. Isorhaphinia Zittel. Sub - cylindrical, pedunculate, with wide cloaca reaching nearly to the base. Skeletal elements large, slightly bent, rod- shaped, inflated at the ends, rarely dichotomously branching. They are associated in bundles, and so interlocked at their extremities as to form an open meshwork. Cretaceous. /. texta (Roemer). Suborder 5. RHIZOMORINA Zittel. Skeletal elements small, composed of four or of three principal rays, or simple and irregulär, with numcrous projecting spines or tubercles ; axial canal simple or branching. Dermal spicules monaxons, tetraxons or similar to those of the skeleton. Chiefly Jurassic, Cretaceous and Recent. Nipterella Hinde. Cambrian. Cnemidiastrum Zittel {Cnemidium, p. p. Goldf.) (Fig. 60). Turbinate or bowl-shaped, with deep cloaca. Walls thick, p^rforated by numerous radial canals disposed in tiers one over another, thus forming vertical fissures which often divide toward the exterior. Skeletal elements irregularly branching, entirely beset with blunt spiny processes. Abundant in the Upper Jurassic 58 COELENTERATA— PORIFEEA PHYLUM II Spongiteiikalk of South Germany, the skeletons being almost invariably replaced by calcite. C. rimulosum Goldf. According to Binde also present in the Carboniferous Limestone of Ireland. Cnemidiastrum stellatum (Goldfuss). Upper Jiirassic Spongiten- kalk ; Hossingen, Württemberg, a, Sponge, ^/^ natural size ; h, Vertical tangential section, showing radial canalsin vertical clefts ; c, Skeletal Clements, 60/j. Fig. 61. ökeleton of Jereica polystoma (Roem.). Upper Cretaceous ; Ahlten, Hanover. 60/. Hyalotragos Zittel. Bowl-, plate- or funnel-shaped, with short peduncle. Depression in summit perforated by the ostia of numerous short canals. Ex- ternal surface finely perforate, or covered by a smooth or wrinkled dermal layer. Skeletal elements irregulär, with numerous branches beset with points, but with few spines. Very abundant in Upper Jurassic Spongitenkalk. H. (Goldfuss). Fig. 62. Chenendoporafungiformis Jjamx. Senonian ; Chatellerault, Touraine. 1/3 natural size. .<^ -. ff "^ ^ ** iV \\^ ^ % y&2 Verruculina auriformis (Roemer). Quadratenkreide; Linden, near Hanover. 2/3 natural size. Platychonia Zittel. Leaf- or ear-shaped, irregularly undulating, covered on both surfaces with fine pores. Skeletal elements resembling those of Hyalo- tragos. Upper Jurassic. P. vagans (Quenstedt). Jereica Zittel (Fig. 61). Sponge cylindrical, turbinate, pyriform or club- ORDER IV SILICISPONGIAE— HEXACTINELLIDA 59 shaped, with short peduncle. Summit truncated or with shallow depression, perforated by the postica of vertical excurreiit caiials. Exterior perforatod by Ostia of the fiiier radial incurrent canals. Skeletal elements root-like, berit, irregularly branching, with numerous short lateral processes. Upper Cretaceous. /. polystoma (Roemer) ; /. punctata (Goldfuss). Chenendopora Lamx. (Fig. 62). Goblet-, funnel- or bowl-shaped, with peduncle. Cloaca deep, perforated by postica of fine canals. Skeletal elements numerously branched and containing branching axial canal. Upper Cretaceous. Verriiculina Zitt. (Fig. 63). Foliate-, funnel-, ear- or bowl-shaped, short- stemmed or sessile. Ostia on the upper surface surrounded by slight, collar- like elevations. Middle (Gault) and Upper Cretaceous. Ämphithelion Zitt. Like the preceding, but with both ostia and postica terminating in bosses. Cretaceous. Other genera : Scytalia, Coelocorypha, Stachyspongia, Pachinion, Seliscothon Zittel ; Megarhiza and Leiochonia Schrammen, etc., in the ^iddle and Upper Cretaceous. Order 4. HEXACTINELLIDA O. Schmidt. (Triaxonia F. E. Schulze.) Süiceous sponges with six-rayed skeletal elements^ the rays being normally disposed in three axes interseding at right angles, and containing axial canals ; elements either detached or fused together so as fo form a lattice-like mesh. Dermal and flesh spicules exceedingly variable in form, but invariably six-rayed. Next to the Lithistida, the Hexactinellida are the most abundant of the fossil siliceous sponges. They are extraordinarily variable in form, and are often anchored by a tuft or "rope" of long, slender, vitreous fibres, or are attached directly by the base. The walls are thin as a rule, and enclose usually a wide cloaca ; the canal-system is consequently much simpler than in the Lithistida., being made up merely of short tubes which penetrate the walls more or less deeply on both sides, and generally end blindly. Sometimes the sponge is entirely composed of thin-walled tubes which twine about one another irregu- larly and produce a System of lacunar interstices (intercanals) of greater or less size. The skeletal elements proper are distinguished by their considerably larger size and simple form from the usually minute, astonishingly variable and delicate flesh-spicules ; the latter, unfortunately, are seldom preserved in the fossil State. The skeletal elements occur detached in the soft parts in the Lyssacina group, or are partially or irregularly cemented together ; in the Dictyonina group, on the other band, they are regularly united in such manner that the rays of proximate elements are all closely applied against one another, and are surrounded by a continuous siliceous envelope. In this way a more or less symmetrical lattice-work with cubical meshes is produced, in which, however, the fusion of juxtaposed elements is indicated in that each ray con- tains two distinct axial canals. The junction of the rays at the central node of each element is usually inflated, but is sometimes sculptured in such manner as to enclose a hollow octahedron (lantern nodes, lychnisks). The exterior of the skeleton is often covered by a dermal layer composed of irregulär hexactins. 60 COELENTERATA— PORIFERA phylum ii in which the externally directed ray has become atrophied ; or a dense siliceous envelope is secreted, in which stellate hexactins with reduced outwardly and inwardly directed rays (stauractins) are embedded in greater or lesser profusion. The Hexactinellida of the present day are distributed chiefly over the greater depths of the ocean beyond the hundred-fathom line (200 to 3000 fathoms). They occur fossil principally in deep-sea deposits, and make their first appear- ance in the Cambrian ; their period of greatest development coincides with Jurassic and Cretaceous time. Suborder 1. LYSSACINA Zittel. ^ Skeletal elements either entirely detached, or only partially and in an irregulär fashion cemented together. Root-tufl offen present. The Lyssacina are poorly adapted for preservation in the fossil state, since the skeletal elements are but rarely cemented together to form a connected framework, and the flesh-sjpicules are invariably destroyed. Notwithstanding, complete sponges composed of large-sized detached hexactins have been found in Paleozoic formations, and also in the Upper Jurassic of Streitberg; and, indeed, the oldest sponges that can be determined with certainty all belong to the Lyssacina. Family 1. Protospongidae Hinde. Thin-walled, sack-, tuhe-like or spherical sponges, with walls composed of a Single layer of cruciform tetraxial spicules (stauractins), arranged so as to form quadrate and suhquadrate meshes. Elements non-fasciculate. The reticulation formed by the larger elements is divided into secondary Squares hy smaller spicules, so that the mesh- work is constituted of several series of Squares. Cambrian and Ordovician. To this family belong the genera Protospongia Salter, and Fhormosella Hinde. Family 2. Dictyospongidae Hall. Usually large, funnel-shaped, cylindrical or prismatic sponges, whose thin walls are frequently diver sified by ridges and prominences. Skeletal framework very regulär, and composed of larger and smaller quadrate meshes situated one within the other. Framework formed by bundles of slender spicules. Chiefly in Devonian (Chemung) and Lower Carboniferous (Keokuk) of North America, and Devonian of Europe. Subfamily 1. Dictyospongiinae Hall and Clarke. Hall and Clarke. Smooth, obconical or subcylindrical sponges devoid of nodes, ridges or other ornamentation ; base furnished with a tuft of long, straight, anchoring spicules. Silurian and Devonian. Hydnoceras Conrad (Fig. 64). Obconical more or less rapidly expanding sponges with eight prism-faces and nodes in horizontal and vertical rows. Base with short tuft of anchoring spicules. Devonian and Carboniferous. Lysactinella Grirty ; Hydriodictya, Prismodictya, Gongylospongia, Botryodictya, Helicodictya, Rhabdosispongia, Ceratodictya, Clathrospongia, Lebedicfya Hall and Clarke. Chemung Group ; New York. ORDER IV SILICISPONGIAE— HEXACTINELLIDA 61 Subfamily 2, Thysanodictyinae Hall and Clarke. Thysanodidya Hall and Clarke. Subcylindrical or tapering Dictyosponges with prominent projecting, rectangularly reticulating spicular bands or lamellae forming series of fenestrated quadrules upon the surface. Base with basal disk or broad obcone. Devonian. Phragmodicfya Hall ; Arystidida, Adoeodiäya, Griphodidya Hall and Clarke. Upper Devonian and Lower Carboniferous. Subfamily 3. Calathospongiinae Hall and Clarke. Calathospongia Hall and Clarke. Stout subcylindrical cups with truncated bases, probably attached by the basal margins ; contracted mesially and more or less expanded at the aperture. Surface without nodes. Carboniferous. Clepsydrospongia Hall and Clarke. Thamnodidya, Cleodidya Hall. Subfamily 4. Physospongiinae Hall and Clarke. Physospongia Hall. Keokuk group. Roemeri- spongia Hall and Clarke. Eifel Devonian. Subfamily 5. Hyphantaeniinae Hall and Clarke. Hyphantaenia (Uphantaenia) Vanuxem. Large, circular and shallow saucer-shaped cups, composed of two series of intersecting spicular straps, one radiating, the other concentric. Chemung Group; New York. Fio. 64. Subfamily 6. Hallodtctyinae Hall and Clarke. cÄ"'''Siemung'G^^^^ . ^ N.Y. Sponge showing foiir rovv.s Hallodidya Hall and Clarke. Admodidya, Crypto- of streng nodes and rineiy reticu- T . -r-r 11 r^^ /-^ "NT ^T ^ lated surface, V? (after Hall and didya Hall. Chemung Group ; New York. ciarke). Subfamily 7. Aglithodictyinae Hall and Clarke. Aglithodidya Hall and Clarke. Chemung Group ; New York. Family 3. Plectospongidae RaufF. Thin-walled tubes with skeleton composed of a regulär frameworh made up of an ascending and approximately ring-like series of spicules ; the latter form redangular to Quadrate, but not very symmetrical meshes. Spicular rays fasciculate. Ordovician and Silurian. Cyathophycus Walcott ; Falaeosaccus, Acanthodidya Hinde. Ordovician. Pledoderma Hinde. Silurian. 62 COELENTEEATA— PORIFEKA PHYLUM II Genera incertae sedis. Pattersonia Miller (Strohilospongia Beecher). In form of large botryoidal clumps. Brachiospongia Marsh. Vase-like sponges with broad inferior margin prolonged into a number of hollow arms. Ordovician of North America. These, together with Ämphispongia Salter, and Astroconia Sollas, from the Silurian of England, represent extinct families of the Lyssacina. Pyritonema M'Coy {Acestra Roem.) Fascicles of long, stout spicules, supposed to be root-tufts. Silurian. Hyalostelia Zitt. (Acanthospongia Young). Skeletal elements relatively large, in the form of regulär hexactins and stellate bodies with reduced vertical ray, and with infiated nodes. Root-tuft composed of elon- gated, slightly bent fibres, sometimes terminating in four recurved rays. Cam- brian to Lower Carboni- ferous ; Great Britain. Holasterella Carter; Spiraäinella (Fig. 65), and Acanthactinella Hinde, are pj^^ qq^ allied genera occurring in ThoUastereiia qracius the Lower Carboniferous ?Sone -^S^f a7 ot Great Britain. f'">- Dermallayerwith ._-r. , ,-^. lused stellate spicules, ThoUastereiia Hmde (r ig. 5/1 (after Hinde). Fig. 65. Spiractinella wrightii (Carter). Carboniferous Limestone ; Sligo, Ireland. A, Normal hexactin. B, Hexactin with forked rays, 0/1 (after Hinde). Fig. 67 Asteractinella expansa Hinde. Carboniferous Limestone ; Dalry, Ayrshire. Skeletal Clement, 5/^ (after Hinde). Astraeospongia menis(ni,s Roemer. Silurian ; Tennessee. A, Sponge, in proüle, 2/3 natural size. B, Upper surface of same. 66), from the Carboniferous, has thin walls composed of a layer of robust, irregularly amalgamated hexactins. As a rule, two of the rays lying in the same plane divide dichotomously from the nodes outward, so as to produce a six-armed instead of a four-armed cross. In Asteractinella Hinde (Fig. 67), all of the rays lying in the same plane divide in two or more branches, thus giving rise to many-rayed, extremely diverse, stellate and corolla-like bodies. Carboniferous ; Ayrshire. Astraeospongia Eoem. (Fig. 68). Thick-walled, depressed, bowl-shaped, Upper surface concave, lower convex, without traces of attachment. Skeleton composed of relatively large, homogeneous, uncemented cruciform spicules ; six of the rays are disposed in the same plane, while the two rays projected at right angles to these are reduced to short, button-like prominences. Common in Silurian of Tennessee and Devonian of the Eifel. ORDER IV SILICISPONGIAE— HEXACTINELLIDA 63 According to Hiiide, Tholiasterella and AsteracMnella constitute a distinct Order (Heteraäinellidae), and AstraeosjJongia is made the type of the order Octadinellidae. These two groups may perhaps best be regarded as aberrant Hexactinellids, in which supernumerary rays are produced by branching. Suborder 2. DICTYONINA Zittel. Skeletal spicules cemented to form, a continuous framework in such a way that every arm of a „ hexaäin is applied to the correspondiiig arm of an adjacent spicule, and hoth rays hecome en- veloped in a common silieeous covering. Root-tuft absent. The Dictyonina are probably de- scendants of the Lyssacina (possi- bly from Froto- spongia- and Dictyo- phyton-Mke forms). They appear first in the Trias, and play a prominent role as rock-builders in the Jurassic and Cretaceous. Their lattice-like skeletons are frequently replaced by calcite, or are dissolved away and merely indicated a ' . h by cavities. The more important fossil forms are divided into the foUowing families. Family 1. Oraticularidae Raiiff. {Eure- tidae p. p., Zittel non Schulze.) Cup-shaped, cylindrical, hranching or flatternd sponges. Spicular nodes solid. External surface without distinct dermal layer, hut protected by a thickening of the outer skeletal layer, and occasionally covered with a delicate web of cemented spicules. Canals simple, blindly terminating in the skeleton. Jurassic. Tremadictyon Zitt. (Fig. 69). Cup-, plate-shaped or cylindrical, with wide cloaca. Canal-openings on both sides in alternating rows. Base nodular; exterior veiled over with delicate net-work of amalgamated hexactins, extend- ing even across canal pores. Skeletal framework with more or less irregulär cubical meshes. Very common in Upper Jurassic. Craticularia Zitt. (Fig. 70). Funnel-shaped, cylindrical or flattened ; Fig. 69. Tremadictyon reticulatum (Goldf.). Upper Jurassic ; ötreitberg, Franconia. a, Sponge, 2/5 natural size ; b, Enlarged portion of outer surface without dermal layer ; c, Portion with well-preserved dermal layer, 3/j ; d, Skeleton, 12/^. Fig. 70. Upper Jurassic ; Craticularia jmradoxa (Münster). Muggendorf, Franconia. a, Sponge, 1/3 natural size ; b, Latticed skeleton, 12/1 ; c, Thickened dermal layer, 64 COELENTERATA— POEIFERA PHYLÜM II simple or branching. Canal - openings on both surfaces either round or elliptical, and regularly distributed in vertical and horizontal rows. Canals short, ending blindly. Jurassic, Cretaceous and Miocene. Sporadopyle Zitt. Cup- to funnel-shaped or conical, occasionally branching. Canal -openings on the outer surface irregularly distributed, or arranged in quincunx ; on the cloacal surface in vertical rows. Upper Jurassic. S. ohliqua (Goldfuss). Sphenaulax Zittel, Ferrucocoelia Etallon. Jurassic. Polyosepia Schrammen. Upper Cretaceous. Family 2. Ooscinoporidae Zittel. Calycoid, beaJcer-like, lobate, branching or stellately convoluted sponges, with thin walls perforated on both sides by numerous canal-openings arranged in alternating rows; canals shori blindly. Framework com- pact, with fine meshes ; dermal layer replaced by thicJcening of the outermost skeletal layer. Spicular nodes solid, more rarely perforate. Cretaceous. Leptophragma Zitt. Beaker - shaped, with root - like attachment. Walls thin, covered on both sides with small canal-openings arranged in alternating rows. Mesh-work very closely woven, spicular nodes solid. Middle and Upper Cretaceous. Pleurostoma'Roem.; Guettardia Mich. ; Balantionella Bchrsimmen. Cretaceous. Coscinopora Goldf. (Fig. 71). Beaker-like, with branching roots. Ostia small, round and in alternating rows. Skeletal elements in part with perforated intersection nodes. Eoot consisting of long siliceous fibres. Dermal layer formed by the thickening and fusion of outermost hexactins. Cretaceous. Fig. 71. Coscinopora infundibuliformis Goldf. Upper Cretaceous ; Coesfeld, West- phalia. a, Complete speciinen, 1/2 natural size ; b, Outer surface, natural size ; c, Same, 3/^ ; d, Skeleton of cup, 12/j ; e, Skeleton of root, i'-^/i. Fainily 3. Staurodermidae Zittel. Turbinate, funnel-shaped or cylindrical, more rarely branching or in clumps. Ostia and postica irregularly distributed, or in alternating rows. Skeletal framework more or less regidar ; intersection nodes thick or octahedrally excavated. The outer or both surfaces of the wall provided with large, stellate spicules (stauractins), which differ from those of the rest of the skeleton, and are either but loosely cemented together or are embedded in a continuous siliceous skin. Jurassic and Cretaceous. Cypellia Zitt. (Fig. 72). Top-shaped, bow-shaped or branching, without root. Canals irregularly distributed, crooked, and branched. Lattice skeleton with irregulär meshes, intersection nodes perforated. Dermal layer 011J)ER IV SILICISPONGIAE— HEXACTINELLIDA 65 composed of large, four-rayed stauractins embedded in a thin, continuous or perforated skin. Very common in Upper Jurassic Spongitenkalk. Stauroderma Zitt. Funnel-shaped or plate-like, with broad and shallow cloaca, into which the large, round postica of short canals open. Inner and outer surfaces provided with dermal layer, in which stellate spicules are embedded with reduced externally and internally directed rays. Upper Jurassic. Casearia Quenst. Cylindrical, with numerous annular constrictions. Cloaca deep, tubiform ; dermal layer relatively thick, and made up of cemented stellate spicules. Upper Jurassic. C. articulata (Goldfuss). Porospongia d'Orb. (Fig. 73). Compressed and expanded, more rarely bulbous or cylindrical. Superior surface pitted with large exhalent apertures of short, blindly terminating cloacae, and covered over with a dense or finely perforate siliceous skin, in which cruciform spicules and regulär hexactins are Cypeläa rugosa (Goldfuss). Upper Jurassic ; Streit- berg, Franconia. a, Spoiige, 1/2 natural size ; b, c, Dermal layer, i^/j. Porospongia impressa (Goldfuss). Upper Jurassic : Muggendorf, Franconia. a, Fragment in natural size b, Dermal layer, 6/^ ; c, Skeleton, 12/^. embedded. Lattice skeleton with cubical meshes ; intersection nodes imper- forate. Upper Jurassic. Family 4. Ventriculitidae Toulmin Smith. ^Fall intricately convoluted ; folds radially disposed, generally vertical in direc- tion. Radial canals ending hlindly. Longitudinal furrows developed along folds of the wall, and either open, or partially covered over with a dermal layer, which is usually formed hy thickening of the outer skeletal layer. Skeletal frameioork with octahedrally perforated nodes. Boots consisting of elongated siliceous fibres united hy transverse hridges and without axial canals. Jurassic and Cretaceous. Pachyteichisma Zittel (Fig. 74). Turbinate or bowl-shaped, with very thick, convoluted wall. Folds separated on outer surface by deeply incised furrows, on inner surface by shallow furrows. Framework extremely regulär. Root and dermal layer absent. Upper Jurassic. Ventriculites Mantell (Fig. 75). Bowl-, plate-, beaker-, funnel-shaped, or cylindrical, with wide cloaca. Wall thin, convoluted ; folds separated on both sides by closely crowded longitudinal furrows. Lattice-work of skeleton more VOL. I F 66 COELENTERATA— PORIFERA PHYLUM II or less regulär; outer layer thickened ; roots present. Common in Middle and Upper Cretaceous. Fia. 74. PachyteicMsma carteri Zittel. Upper Jurassic ; Hohenpölz, Franconia. a, Sponge, 1/2 natural size ; b, Skeleton, 12/j. Fig. 75. VentricuUtes striatvs Smith. Qnadratenkreide ; Linden, near Hanover. a, Sponge, 1/2 natural size ; b, Transverse section, 1/1 ; c, Skeleton, 12/j. Fio. 76. CoeloptycTinim agaricoides Goldf. Upper Cretaceous ; Vordorf, near Brunswick. A, Top view. li, Profile. C, Under surface, 2/3 natural size. D, Skeleton, 60/j, Schizorhabdus, Bhizopoterion, PolyllasUdium Zittel ; Sporadoscinia Pomel ; 1 ER IV SILICISPONGIAE— HEXACTINELLIDA 67 Lepidospongia Roemer; Leiostracosia, Plectodermatium, Microblastidium Schram- men, etc. Cretaceous. Family 5, Ooeloptychidae ZitteL Umbel- or mushroom-shaped, with stalk. Wall thin, deeply folded. Convolutions radialhj arranged, hecoming furcate toward periphery of umbel, and exposed on lower surface. Marginal and ^^ Upper surface enveloped with porous dermal layer entirely covering thefolds. Ostia only on under side of umbel, situated on backsof thefolds. Frame- work very regulär ; inter- sedion nodes odahedral, perforated ; rays ofhex- actins provided with slender, thorny processes. Coeloptychium Goldf. (Fig. 76), oc- curring in the Upper Cretaceous of Northern Germany, England, and Southern Russia, is the solitary genus. Fig. 77. Flocoscyphia pertusa Gein. I Greensand (Cenomanian) «.'Fragment in\natural size ;l&, Dermal layer, live times enlarged of interior, 12/^1; d, Outward portion of skeleton, 12/j Banowitz, Hungary. Skeleton Family 6. Maeandrospongidae ZitteL Sponge body consisting of thin-walled, intricately labyrinthine, and partially amalgamated tubes or foliae, which form tuberous, pyriform, beaker-shaped, or bush- ß like branching Stocks. Between the tubes are cavities and interstices of considerable size, which constitute the so- called intercanalicular System. Four canals faintly developed. Der- mal layer absent, or re- presented by a continuous silicious superficial skin. Abundant in the Cre- taceous, and also re- presented by numerous recent genera. Flocoscyphia Reuss (Fig. 77). Clump-like or bulbous Stocks con- sisting of labyrinthic, anastomosing tubes or foliae. Walls of tubes thin, perforated by numerous small Ostia. Latticed skeleton, intersection nodes solid or perforate. Cre- taceous. Becksia Schlüter (Fig. 78). The thin walls of the shallow, beaker-like Fig. 78. Becksia sokdandi Schlüt. Quadratenkreide ; Coesfeld, Westphalia. A, Sponge body, 1/2 natural size ; 0, Ostia of radial canals ; /, Hollow, root-like processes of wall. B, Skeleton, 50/j. COELENTEKATA— POEIFEKA PHYLUM II sponge are composed of vertical tubes having a radial disposition and fused with one another along the sides. Between the tubes are large interstices ; near the base the tubes develop hollow, spinous processes. Lattice skeleton very i regulär, exactly similar to Coeloptijchium. Upper Cretaceous ; Westphalia. Tremaholites Zitt. ; Etheridgia Täte ; Zittelisjjongia SinzofF, etc. Upper Cretaceous. Camerospongia d'Orb. (Fig. 79). Globular, sub-globular, or pyriform. FlG Camerospongia fungiformis (Goldfuss). Plänerkalk ; Oppeln, Silesia. Natural size. Fig. 80. Cystispongia hursa Quenst. Cuvieri-Pläner (Turonian); Salz- gitter, Hanover. a, Sponge, natiiral size ; h, Dermal layer with underlying skeletal framework ; c, Skeleton, 12/^. Upper half of the body enveloped by smooth siliceous skin, and with large circular depression on the summit ; lower half marked b}^ undulating ridges and furrows, and passing gradually into a stem. Interior of sponge body consists of thin-walled, labyrinthous tubes. Upper Cretaceous. Cystispongia Eoem. (Fig. 80). Like the preceding, but with dense siliceous skin punctured by large, irregularly shaped apertures, uniforraly enveloping the whole sponge body. Body composed entirely of tubes. Cretaceous and still living. Subclass 4. CALCISPONGIAE. Calcareous Sponges. Skeleton composed of calcareous spicules of three-rayed, four-rayed, or uniaxial The external form of the Calcisponges is quite as variable as that of the siliceous sponges, and reminds one particularly of the Lithistida. Like the Lithistids, too, the thick-walled Leucones and Pharetrones have a canal-system consisting of a central cavity into which radial excurrent canals conduct ; while the numerous tributaries of the latter end in ciliated Chambers which are fed by fine incurrent canals. In the Sycones the wall is perforated by simple radial tubes, but in the thin-walled Ascones it is pierced by mere holes. The calcareous skeletal Clements lie free in the soft parts, sometimes forming but a single layer disposed in the same plane {Ascones) ; sometimes their disposition is more or less distinctly radial, following the canal courses (Sycones) ; sometimes they are irregularly crowded together (Leucones) ; and ORDER I CALCISPONGIAE— PHARETRONES 69 ^^^^^^^^ sometimes they are closely opposed iii the form of solid anastomosing fibres (Pharetrones). Regulär triaxial spicules are of the most common occurrence, next monaxial spicules, sharpened on both sides, and more rarely four-rayed spicules. Ovving to the ready solubility of the skeletal Clements in calcareous sponges, they are usually but poorly pre- served in the fossil state, and are ill-adapted for micro- scopical investigation. The three-rayed and rod-shaped spicules which are united in fibres are seldom distinctly recognisable as such, since, as a rule, they are either wholly or partially dissolved, and are converted into homogeneous or cry stalline fibres of calcite (Fig. 84) ; in these minute threads of calcite may be seen radiating in all directions from numerous centres of crystallisation. Sometimes such calcareous skeletons afterwards become silicified. It is clear, therefore, that the present chemical composition of a fossil sponge furnishes us no clue in regard to its original character, since during the process of fossilisation a Triaxial skeletal elements of a Recent Ascon, ^/i. Fio. 82, Fibres of a Pharetrone, composed of three- Solid fibres of fossil calcar- Fibres of fossil calcareous rayed spicules. Feronidella cylindrica (Gold- eous sponge with partially sponge altered by crystallisa- fuss). Upper Jurassic, 40/j. preserved spicules, ^o/^. tion, 40/^. siliceous skeleton may become converted into a calcareous, and a calcareous into a siliceous. Of the four Orders of calcareous sponges — Pharetrones, Sycones, Ascones, and Leucones — only the first two are of practical importance to the paleontologist, traces of the others being either wanting or extremely fragmentary. Order 1. PHARETRONES Zittel. ^rall thick ; canal System like that of the Lithistida, though sometimes indistinct and apparently absent. Spicules arranged in solid anastomosing fibres ; a smooth or corrugated dermal layer frequently present. Devonian to Cretaceous ; unknown in Tertiary and Recent. Eudea Lamx. Cylindrical or club-shaped, usually simple, rarely branching. Cloaca narrow, tubiform, extending to the base, and terminating above in a round osculum. Dermal layer smooth, perforated by ostia of short canals. Triassic and Jurassic. E, clavata Lamx. Peronidella Zitt. (Peronella Zitt. non Gray ; Siphonocoelia, Polycoelia From.), (Figs. 85, 86). Thick - walled, cylindrical, simple or branching. Cloaca tubiform, extending to the base ; base sometimes covered by a dense dermal layer. The rest of the exterior finely perforate. A distinct canal-system 70 COELENTERATA— PORIFERA PHYLÜM II absent. The coarse, anastomosing skeletal fibres composed of closely packed three-rayed and one-rayed spicules. Sparse in Devonian (P. constricta Sandb.) ; common in Trias, Jurassic, and Cretaceous. Fig. 85. Peronidella cylin- drica (Münst.). Upper Jura. ; Miig- gendorf. x 1/2. Fio. 86. Peronidella diimosa (From.). Hils ; Berklingen, Bruns- wick. Natural size. Fig. 87. Corynella quenstedti Zitt. Coral-Rag; Natt- heim. 0, Sponge, natural size ; b, Skeletal fibres, 4/1. Fig. 88. Eusiphonella bronni (Münst.). Coral-Rag ; Nattheim. Natural size. Oculospongia tu- hulifera (Goldf.). Kreidetuflf; Mae- stricht. size. Fig. 90. Eusiphonella Zitt. (Fig. 88). Similar to preceding, but thin-walled, with broad cloaca extending to the base, into which con- duct radial canals ar- ranged in vertical rows. External surface per- forate. Upper Jurassic. Corynella Zitt. (Fig. 87). Knob-like, cylindri- cal, or top-shaped, thick- walled, simple, or com- Natural (Q^eSr €0?^^ P^'^*®' ^^^^^^ ^^^^^^ " Nattheim. Natural size. shaped, shallow, terminat- ing below ^in a series of vertical branching tubes ; exhalent aperture often surrounded with radially diverging furrows. Ostia conducting into numer- ously branching radial canals, which unite again in larger excurrent canals, and open into the cloaca. Common in Trias, Jur- assic, and Cretaceous. Stellispongia d'Orb. (Fig. 90). Usually com- posite Stocks made up of hemispherical, or short pear-shaped perSOnS, with Upper surface, natural size. base enveloped by com- pact dermal layer. Summit dome-shaped, with shallow cloaca surrounded by radial furrows ; radial and vertical canals terminating along sides and basis of Hils Fig. 91. Elasmostoma acntimargo Roem. Berklingen, Brunswick. Fig. 92. RhapMdo nema farringdonense (Sharpe). Lower Cretaceous (Aptian) ; Farringdon, Berkshire. 2/0 natural size. ORDER II CALCISPONGIAE— SYCONES 71 cloaca. Skeleton constituted of short, bluiit, and bent uniaxial, and also of three- and four-rayed spicules. Triassic and Jurassic. Holcospongia Hinde. Jurassic and Cretaceous. Sestromostella Zittel. Trias to Cretaceous. Synopella Zittel. Cretaceous. Oculospongia (Fig. 89) and Diplostoma From. Cretaceous. Elasmostoma From. (Fig. 91). Foliate-, ear-, or funnel-shaped. Upper {i.e. inner) surface covered with smooth dermal layer, in which large shallow oscula are situated ; under surface cribriform. Cretaceous. Bhaphidonema Hinde (Fig. 92). Beaker-, funnel-, or twisted leaf-shaped. Inner or upper surface smooth, with very small oscula or pores. Outer surface rough, cribriform. Canal-system indistinct. Trias, Jurassic, Cretaceous. Pachytylodia Zitt. Funnel-shaped, thick-walled ; base with smooth dermal layer; oscula present here, but absent on other parts of the exterior. Skeleton composed of very coarse, anastomosing fibres. Cretaceous. P. infmidihulifm'mis (Goldfuss). Order 2. SYCONES Haeckel. Walls traversed hy simple canals disposed radially with reference to the cloaca and opening into it. Skeletal elements very regulaiiy arranged. Mostly small delicate forms inhabiting shallow water. Protosycon Zitt., from the Upper Jurassic of Streitberg, is a small, cylin- drical, or conical form agreeing with living Sycons in the arrangement of its radial canals. To the Sycons, Rauff assigns also the calcareous sponge Sphinctozoa described by Steinmann (Jahrb. f. Mine- a ralog. 1882, IL p. 139), which is distinguished from all other Calcisponges by having a most remarkable segmentation, such as occurs in the Lithistid genus Casearia. The oldest Sycons are Sollasia, Amhlysiphonella and Sebargasia Steinm., from the Carboniferous Limestone of Asturias. In the Triassic of St. Cassian and Seelandalp, near Schluderbach in Tyrol, are found Colospongia Laube, Thaimastocoelia and Cryptocoelia Steinmann. Thalamopora Roemer and Barroisia Steinm., occur in the Lower and Middle Cretaceous. Barroisia (Fentriculites Zitt. non Defr. ; Sphaerocoelia Steinm.) (Fig. 93). Occurs sometimes as simple, cylindrical, or clavate individuals, and again in the form of bushy Stocks. Outer surface frequently constricted, summit arched, with osculum in the centre, cloaca tubiform. The cylindrical indi- viduals are composed of thin-walled, hemispherical, or compressed segments, which are so arranged that the roof of one segment serves also as the floor of the next following. The wall is every where perforated by simple radial canals, Fig. 93. Barroisia anastomans (Mantell). Aptian ; Farringdon, Berk- shire. A, ßush-like colony, one branch sliced open ; natural size. B, Individual cut through obliquely, 5/2 ; a, Junction of two Segments ; 6, Cloaca ; 0, Osculum ; d, Radial canals. C, D, Three-rayed skeletal spicules, 3«/i and 72/i (after Steinmann). 72 GOELENTEKATA— POEIFERA phylum ii and is made up of fibres composed of three-rayed spicules. B. helvetica (Lor.). Aptian ; La Presta, Switzerland. Appendix to Sponges. Incertae seclis. Family. Receptaculitidae Koemer.i This Singular group which ranges througbout the Ordovician, Silurian, and Devonian Systems, consists of globular, cup-, or platter - shaped bodies containing a central cavity, and whose wall is composed of elements arranged in quincunxial order. The substance of the wall is thought by Hinde to have been siliceous ; calcareous according to Kauff ; aragonite according to Gümbel ; calcite or chitinous according to Billings, either aragonite or chitinous in the opinion of Girty. The elements lying on the outer or under side of the wall have been usually described as consisting of small rhomboidal plates having four transverse rays disposed crosswise, and one inwardly directed ray ; but Girty has found evidence that the spicular summit plates are infiltrations of the rhombic pits of the outer surface, and the radial pillars or spicules are infiltrations filling radial tubes. The systematic position of these problematic fossils is wholly conjectural. Gümbel assigns them to the calcareous algae (Dadyloporidae), and others to the Foraminifera and Sponges. Hinde has referred them to the HexadinelUda, but the observations of Eauff and Girty as to the original calcareous and chitinous composition of the wall disprove this inference. BeceptacuUtes Defrance. Spherical or pyriform bodies, with a central closed cavity. Ordovician to Carbon iferous. Europe, America and Australia. Ischadites Murchison (Dictyocrinites Conrad ; Didyocrinus Hall). Conical or ovate bodies, inclosing a central cavity, with a small summit aperture and lacking an inner layer. Ordovician to Devonian; Europe and America. Here are also referred Cydocrinus Eichwald ; Pasceolus Billings ; Polygono- sphaerites Eoemer; Cerionites Meek and Worthen ; Lepidolites and Anomalospongia (Anomaloides) Ulrich. Range and Distribution of Fossil Sponges. The phylogeny of the Myxospongiae, Ceratospongiae and a part of the Silid- spongiae, owing to their perishable Organisation, remains involved in doubt. 1 Salter, J. TF., CanadianOrganic Remains, Dec. 1, 1859.— Hall, J., Pal. N. Y., vol. i., 1847 ; Geological Report of Wisconsin, 1862 ; Sixteentli Rept. N. Y. State Cabinet Nat. Hist., 1863 ; Twelfth Rept. State Geologist of Indiana, 1883 ; Palaeontology of New York, vol. iii., 1859 ; Eleventh Rept. State Geologist of Indiana, 1882 ; Second Ann. Rep. N. Y. State Geologist, 1883 ; Palaeontology of New York, vol. vi., 1887. — Ulrich, E. 0., Jour. Cincinnati Soc. Nat. Hist., vol. i., 1871; vol. ii., 1879. — Owen, D. />., Geol. Report Iowa, Wisconsin and Illinois, 1844; Geol. Survey of Wisconsin, Iowa and Minnesota, 1852. — Billings, E., Palaeozoic Fossils, vol. i., 1865 ; Canadian Naturalist and Geologist, second ser., vol. ii., 1865 — Meek and Worthen, Geol. Survey of Illinois, vol. iii., 1868. — Gümbel, G. W., Abhandl. der k. bayr. Akad. Wissensch., vol. xii., 1875. — Roemer, F., Letliaea Palaeozoica, 1880. — Hinde, G. J., Quart. Jour. Geol. Soc. London, vol. xl., \8M.— James, J. F., Jour. Cincinnati Soc. Nat. Hist., vol. viii., 1885 ; vol. xiv., ISdl.— Walcott, C. 1)., Mon. U. S. Geol. Surv., vol. viii., 1884:.— Whüßeld, R. P., Geology of Wisconsin, vol. iv., 1884.— Rauff, H., Zeitschr. deutsch, geol. Gesellsch., vol. xi., 1888. — Nicholson and Lydekker, Manual of Palaeontology, vol. ii., 1889 — Winchell and Schuchert, Geol. of Minnesota, vol. iii., pt. 1, Pal. 1895. — Ulrich, E. 0., ibid., p. 68— Girty, G. //., Fourteentli Ann. Rept. N. Y. State Geologist for 1894, 1895. — Weller, S., Geol. Survey of New Jersey, Rept. on Pal., vol. iii., 1903. I RANGE OF FOSSIL SFONGES 73 Nevertheless, isolated spicules prove the existence of Monactinellids and Tetractinellids in Paleozoic seas ; while in the Trias, Jura and Cretaceous these forms become important rock-builders, and play an active part in the formation of hornstone, chalcedony and flint. In the Tertiary, spicules refer- able to existing genera are common. The former distribution of the three best preserved sponge groups — the Lithistids, Hexactinellids, and Calcisponges — is noteworthy. The living repre- sentatives of the first two Orders inhabit deep or moderately deep water, while the calcareous sponges predominate in shallow waters bordering the coast. And hence, since fossil Calcisponges likewise occur almost entirely in marly, clayey, or sandy strata of undoubted littoral origin, and are absent in lime- stones where Lithistids and Hexactinellids predominate, it is piain that the distribution of both fossil and Recent sponges has been occasioned by like physical conditions. In the Cambrian occur the Lithistid genera Archaeoscyphia and Nipterella, and in the Ordovician and Silurian of Europe and North America are found a number of Tetracladina (Aulocopium) and Eutaxidadina forms {Astylospongia, Palaeomanon, Hindia), together with a few Rhizomorina. In the Carboniferous ßhizomorina and Megamorina are sparsely represented ; but in the Upper Jurassic, and especially in the Spongitenkalk of Franconia, Swabia, Switzer- land, and the Krakau district, the Lithistids exhibit an astonishing develop- ment, and occasionally form thick beds. They occur only sparingly in the Lower Cretaceous, but are abundant in the Pläner, Greensand and Upper Cretaceous of Northern Germany, Bohemia, Poland, Galicia, Southern Russia, England and France. The Tertiary being nearly everywhere made up of shallow-water formations, the absence of Lithistids and Hexactinellids is not surprising. They persist locally, however, as in the Upper Miocene of Bologna and in the Province of Oran in Northern Africa. The ränge of the Hexadinellida is in every respect similar to that of the Lithisiida. Beginning in the Upper Cambrian, they are represented in the Ordovician and Silurian by peculiarly modified Lyssacina forms (Protospongia, Phormosella, Cyathophycus, Palaeosaccus, Pledoderma, Pattersonia, Brachiospongia, Didyophyton, Astraeospongia). The same group continues also through the Devonian, where Didyophyton and its associates are conspicuous for their wide- spread distribution in North America. A few aberrant Lyssacina, which Hinde designates as Heteradinellidae, are found in the Carboniferous. During the Mesozoic and Cenozoic eras the distribution of the Hexadinellida is nearly identical with that of the Lithistida; although here and there beds occur which are charged principally with Hexactinellids, and others chiefly with Lithistids. Very different conditions are presented by the Calcisponges, among which only the Pharetrones and Sycons are of geological importance. The oldest calcareous sponges occur very sparsely indeed in the Middle Devonian and Carboniferous Limestone. They appear in considerable diversity in the Alpine Trias (St. Cassian and Seelandalp), but outside the Alps are almost wholly absent. In the Jurassic they occur in marly beds of the Dogger (Ranville, Swabia), and also in certain facies of the Malm (Terrain ä Chailles, Coral-Rag of Nattheim, Sontheim, etc.) in Southern Germany and Switzerland. The Lower Cretaceous, particularly the Neocomian of Brunswick, the Swiss Jura, and the Paris Basin, as well as the Aptian of La Presta, near 74 COELENTEKATA— CNIDARIA phylum ii Neuchätel, and Farringdon, Berkshire ; and also the Middle Cretaceous (Ceno- manian) of Essen, Le Mans, and Havre, are characterised by an abundance of well-preserved Pharetrones, and a lesser number of Sphinctozooid Sycons. In the Tertiary, however, both groups are wanting, although the existence of calcareous sponges is still indicated by occasional detached triactins. The Pharetrones apparently become extinct at the close of the Cretaceous. SuBPHYLUM II. Cnidaria. The Cnidaria or Nematophora have a radially symmetrical body, and a terminal mouth-opening surrounded by fleshy tentacles. In the ectoderm (sometimes also in the entoderm) cnidoblasts are common, from the contents of which thread-cells (nematocysts) filled with an urticating fluid and containing a hollow, spirally coiled thread, are developed. Each cnidoblast possesses a fine superficial process (cnidocil), which is very sensitive to mechanical Stimuli, The polyp wall typically consists of three layers : an outer ectoderm, an inner endoderm, and a middle mesogloea. The mesogloea is sometimes entirely absent, but the ectoderm and entoderm are strongly developed. The ectoderm frequently secretes a calcareous or horny skeleton, and both ectoderm and entoderm are concerned in the production of muscles and nerves. The sexual Organs are the product of the entoderm. The Cnidaria are divided into two classes : Anthozoa and Hydrozoa. The latter are undoubtedly the more primitive group, but it will be convenient to treat of the Anthozoa first in the present work. Olass 1. ANTHOZOA = A0TINOZOA. Coral Polyps.i Usually sessikj cylindrical polyps, possessing a mouth surrounded by tentacles, Oesophagus, and gastrovascular cavity. The latter is divided by numerous vertical partitions (mesenteric folds) into a System of radially disposed pouches. A calcareous or horny skeleton is frequently developed. Simple or forming colonies. The simple polyp zooids have the form of a cylindrical or conical tube at the distal end of which is situated a muscular disk perforated centrally by ^ Literature : Milne Edwards, H., et Haime, J., Histoire naturelle des coralliaires, 3 vols. and atlas. Paris, 1857-60. — Idem, Monographie des polypiers fossiles des terrains paleozoiques. Arch. du Museum, Paris, vol. v., 1851. — Idem, Monograph of the British Fossil Corals. Palaeontogr. Soc, 1849-64. — Fromentel, E. de, Introduction a l'etude des polypiers fossiles. Paris, 1858-61. — Idem, Paleontologie frangaise ; 1861 and later. — Reuss, A. E., Articles in Sitzber. Akad. Wiss. Wien, 1859, 1864, 1865, 1870 ; also Denkschr. vols. vii., xxiii., xxviii., xxix., xxxi., xxxiii. — Duncan, P. M., British Fossil Corals, 2d ser. Palaeontogr. Soc. 1865-69, and 1872. — Idem, Revision of the Families and Genera of the Sclerodermic Zoantharia or Madreporaria. Journ. Linn. Soc. Zoology, 1885, vol. xviii. — Kohy, F., Monographie des polypiers jurassiques de la Suisse. Abhandl. Schweiz. Pal. Ges., 1880-94, vols. vii.-xxii. — Pratz, E., Ueber Septalstructur. Palaeontogr. 1882, vol. xxix. — Koch, G. von, Die ungeschlechtliche Vermehrung der paläozoischen Korallen. — Ibid., 1883, vol. xxix. — Quenstedt, F. A., Petrefactenkunde Deutschlands, 1889, vol. vii. — Kohy, F., Monographie des poly- piers cretaces de la Suisse. Abhandl. Schweiz. Pal. Ges. 1896-98, vols. xxii.-xxiv. — Ogüvie-Gordon, Maria M., Korallen der Stramberger Schichten. Palaeontographica, Supp. II., 1897. — Idem, Systematic Study of Madreporarian Types of Corals. Phil. Trans. Roy. Soc. London, 1897, ser. B, vol. clxxxvii. — Gregory, J. W., I^he Corals, Jurassic Fauna of Cutch. Palaeontol. Indica, 1900, ser. 2, vol. ix., pt. 2. — Vaughan, T. Wayland, Eocene and Lower Oligocene Coral Faunas of the United States. Mon. xxxix, U.S. Geol. Survey, 1900. — Ide7n, Critical Review of the Literature on the simple Genera of the Madreporaria Fungida. Proc. U.S. Nat. Mus., 1905, vol. xxviii. — Buerdeii, J. E., West ludian Madreporarian polyps. Mem. Nat. Acad., 1902, vol. viii, — Idem, 5S I ANTHOZOA 75 ^^|l|e slit-like or oval fissure of the mouth. The oral disk is furnished with a ring of tentacles round its margin, and opeiis into a membranous oesophageal tube conducting into the gastric cavity. The outer covering of the body, the parts of which are designated as wall, oral disk, and pedal disk, are constituted of ectoderm and entoderm, between which is a thin layer of mesoderm (mesogloea). Six, eight, or more radially disposed vertical partitions (mesenteries), (Figs. 94, 95), projecting inwardly from the body-wall, divide the gastric cavity into a series of radiating compartments (mesenteric pouches). The mesenteries are continuous upwardly with the hollow, muscular tentacles ; while the generative Organs are attached to their faces near the lower end of the body. The mesenteries are »covered on both sides with muscular tissues, and bear mesen- teric filaments on their curled inner edges. On one side of the mesenteries the muscle fibres are transversely directed, on the other longitudinally. The longitudinal System is usually considerably folded and thickened ; and the disposition of these muscular portions is of great importance from a systematic standpoint, since it reveals the bilateral symmetry of many Anthozoans, and enables one readily to identify the antimeres. If a polyp individual be cut in two by a plane passing through the longer axis of the mouth-opening, then, in the Octocoralla (Fig 94), the mesenteries of the right half will have all the muscular thickenings disposed on the right-hand side, and those of the left on the left-hand side. In the Hexacoralla (Fig. 95) the mesenteries are grouped in pairs, with the muscular thickenings of any pair facing each other. Two pairs, however (those corresponding with the opposite extremities of the longitudinal mouth), form often an exception to this rule, since these have the muscular thickenings placed on opposite sides. These are called the directive mesenteries, and serve to indicate the longitudinal axis of the body. Only a few Anthozoa have permanently soft bodies ; the majority secreting calcareous, horny, or partly horny and partly calcareous structures, termed the skeleton or corallum. The simplest form of corallum is that composed of microscopic, round, cylindrical, acerate, or tuberculated spicules of carbonate Fig. 94. Diagrammatic section of the soft parts of an Octo- coralla (Alcyoninm). x, Oeso- phagus ; 1, 2, 3, 4, Mesen- teries of the left side (after R. Hertwig). Fig. 95. Diagrammatic section of the soft parts of a Hexacoralla. In the upper half (above the line a — h) the section passes through Oeso- phagus s ; in the lower half, beneath the same. Corallum indicated by heavy lines. r, directive mesenteries. Morphology of the Madreporaria. A series of papers in Ann. and Mag. Nat. Hist., ser. 7, vols. ix., X., xi., xvii.,xviii. (1902-1906), and Biol. Bull.,vols. vii. and ix. (1904-1905).— /rfew, Recent Results on the Morpliology and Development of Coral Polyps. Smithsonian Miscellaneous Collections, Quart. Lss., 1904, vol. xlvii. — Felix, J., Die Anthozoen der Gosan Schichten in den Ostalpen. Palaeontographica, 1903, vol. xlix. Nuinerous other papers, especially in Zeitsch. deutsch, geol. Gesellsch. — Caimthers, R. G., The primary Septal Plan of the Rugosa. Ann. and Mag. Nat. Hist. 1906, ser. 7, vol. xviii. — Gordon, O. E., Studies on early Stages in Paleozoic corals. Am. Jour. Sei., 1906, vol. xxi. — Bnrwn, T. 0., Studies on the Morphology and Development of certain Rugose Corals. Ann. N.Y. Acad. Sei., 1909, vol. xix. 76 COELENTERATA— CNIDARIA PHYLÜM II of lime, which are developed in great quantities and remain detached in the soft parts (many .Alcijonaria). In a number of forms (Corallium, Mopsea, Tubipora) the spicules are firmly cemented together hy means of a calcareous or horny connective substance, in such a manner as to form tubes (Tubipora), or, when the secretion takes place chiefly at the base, a sclerobase, or axis. Surrounding the axis is the soft coenosarc in which the polyps of the colony are embedded (Fig. 96). In some cases the sclerobase is composed entirely of horny matter without admixture of calcareous secretions. In the so-called "stone corals" (Fig. 97) a consistent calcareous skeleton is formed by the oiiter surface of the ectoderm. At the base of the polyp between each pair of mesenteries, the infolded ectoderm secretes small, round, cgi^al or irregulär calcareous bodies (sderites) ; these are opposed against one another in radial directions, and as others are successively laid down on top of them, upright Corallium rubrum Lam. (after Lacaze-Duthiers). Brauch of red coral of commerce laid open along the axis, and showing three polyps in section embedded in fleshy coenosarc. Fig. 97. Astroides calycularis (Lamx.). Mediterran ean (alter Lacaze-Duthiers). Enlarged longitudinal section of polyp with calcareous skeleton. te, Tentacles ; oe, Oesophagus ; me, Mesentery ; loc, Mesenteric pouches ; coe, Coenosarc ; spt, Septum ; col, Columella. partitions or septa are built up. Early in development also, after fixation of the larva, the basal plate becomes calcified, owing to the secretion by the outer surface of the ectoderm of numerous minute calcareous granules (calicoblasts). The septa, however, grow considerably above the base, and become lodged in the vertical interspaces between the mesenteries. In the same manner, within the soft body-wall, a calcareous secretion may take place, binding the outer borders of the septa together, and known as the wall or theca. Both septa and theca are composed of minute, densely crowded calcareous bodies, in which delicate calcareous fibres may be seen radiating in all directions from a central dark space. And since all the calcareous bodies forming the septa have a radial disposition, the calcification-centres as seen in transverse sections form a dark, mostly interrupted and occasionally jagged median line, from which bundles of minute fibres radiate outward in all directions. Similar calcification- centres may also be found in the theca. Sometimes the median dark line is uninterrupted and divides the septum into two separate lamellae. i. ANTHOZOA 77 The interstices between the sclerites forming the septa are either com- pletely filled with carbonate of lime (Aporosa), or there remain larger or smaller porous Spaces (Ferforata) ; in many cases, in fact, the septa are represented by a loose network of sclerites piled up vertically, or merely by vertically directed spines. The number of septa and of tentacles is either equal to the number of pairs of mesenteries (when only entocoelic septa are present), or double that of the pairs of mesenteries (when both exocoelic and entocoelic septa and tentacles are present), and is somewhat uniform throughout species, genera and higher groups. The number, width and mode of formation of the septa furnish important systematic characters. As the locus of the origin of the septa succeeding the primaries may conform to one of several plans, this character is used in determining the major groups. The upper edges of the septa are sometimes smooth, sometimes serrated or granulated ; and they extend from the central depression to or through the walls of the theca, either obliquely or in a curved line. This open, central depres- sion, formed by the superior edges of the septa, is known as the calice or calyx. The sides of the septa are rarely smooth, but are commoniy granulated or furnished with rows of small prominences ; occasion- ally they are provided with well-marked vertical cross-bars (carinae). When the projections on the sides of the septa are in the form of conical or cylindrical trans- verse bars, they are termed synapticulae. Frequently the synapticulae of two adjacent septa become joined together ; sometimes whole rows of them are fused together to form perpendicular bars, thus greatly strengthening the septal framework. In some corals (athecalia) the development of synapticulae is such as to render an outer wall superfluous. With the upward growth of the polyp, the theca gradually becomes elevated, and its lower portions, as their occupation by the soft parts ceases, may be partitioned ofF by numerous horizontal or oblique calcareous plates which bridge over the interseptal Spaces. These structures are known as dissepiments and tabulae. The tabulae are often nothing but highly developed dissepiments, being distinguished from the latter merely by the fact that they extend across between the septa at the same level; sometimes they are perfectly horizontal, sometimes they are arched or funnel-shaped (Fig. 98), and sometimes incomplete. Dissepiments and tabulae are most strongly developed in cylindrical forms, and frequently fill the included space within the theca with a vesicular or cellular tissue. When a number or when all of the septa are produced as far as the centre of the calice, their inner edges may become twisted so as to form an axial structure, known as a pseiidocolmnella. Sometimes, however, a true columella is present ; this may be either a compact, styliform or foliaceous structure, or Lithostrotion martini B. and H. Longitudinal sec- tion showing tabulae. Fig. 99. CaryophylUa cyathus Sol. Corallum split open longi- tudinally ; true columella in the centre, surrounded by a cycle of pali. 18 COELENTERATA— CNIDARIA phylum ii may be composed of a bündle of styliform or twisted rods (Fig. 99), or of thin lamellae. It extends from the floor of the visceral Chamber to the bottoni of the calice, into which it projects for a greater or less distance. The structures known as pali are narrow vertical plates which are inserted between the colu- mella and the inner ends of the septa in one or more cycles (Fig. 99). The outer wall or theca is often formed by the secretion of a particular ring-like fold of the ectoderm, and is constituted of distinct sclerites, having separate calcification-centres, and connecting the outer borders of the septa (euthecalia). In many cases the peripheral edges of the septa become thickened and laterally fused to form a spurious theca (pseudothecalia) ; and occasionally the dissepiments lying in a certain zone become united so as to form an inner wall within the true theca. The epitheca is a usually smooth, sometimes corru- gated, superficial calcareous investment, which, according to Koch, is merely a Prolongation of the basal plate, and is secreted by the outer surface of the ectoderm, which is reflected over the top of the corallum. The epitheca is deposited either directly upon the septa, or upon the theca, or, when the septa are produced outwards so as to form exothecal lamellae or ribs (costae), the theca and epitheca are separated. Exothecal lamellae, not corresponding in Position to the septa, are called pseudocostae or rugae. New individuals or colonies commonly originate by sexual reproduction. Following fertilisation and segmentation of the ova, ciliated larvae are born, which swim about for a time, become fixed, and develop into simple polyp individuals. Vegetative or asexual increase by two sharply defined processes, namely, hudding (or gemmation) and ^jssiow, assumes a great importance among Anthozoans, resulting in the production of colonies or Stocks, often of large size and exceeding complexity of form. New corallites are produced either within or without the calice of the parent polyp. In extra-calicinal gemmation the buds are thrown out either from the sides of the polyp (lateral gemmation), or are formed in the common calcareous matrix which unites the various corallites of a colony {coenenchymal and costal gemmation). In both cases the new corallites may diverge from one another, being attached to the parent corallum only at the base, or they may grow up closely opposed to the latter and to one another, so that the thecae are in contact on all sides. In this way branched, dendroid or massive and knob-like '("astraeiform") Compound coralla are formed. A less common mode of increase is by hasal or stolonal gemmation. In this process the wall of the original polyp sends out creeping prolongations (stolons) or basal expansions, from which new corallites arise. In calicinal gemmation buds are produced within the calice of the parent corallite, according to one or the other of the following methods : either certain particular septa become enlarged and pro- duced so as finally to enclose a new calicinal disk {septal gemmation) ; or tabulae are produced upwards in the form of pockets, from which new corallites are developed (tahular gemmation). In both septal and tabular gemmation, a portion of the parent corallite including a part of the original wall is concerned in the formation of buds ; while the septa or modified tabulae are converted into portions of the new thecae, from which new septa then begin to grow inwards towards the centre. A peculiar kind of calicinal gemmation is that known as rejuvenescence. In this method only one bud is formed within the parent calice, but it enlarges aintil it completelj ßlls the latter. By the indefinite repetition of this process, a CLASS I ANTHOZOA 79 corallum is formed, consisting of a succession of cups placed one within the other, of which only the youngest and uppermost is occupied by the living animal. The beginning of reproduction by fission is marked by an elongation or distortion of the parent calice, accompanied by the contraction of the wall at opposite points along the margin. The constriction may proceed until it divides the oral disk into two halves; or two opposite septa may unite to form a new theca. By this method branching, massive or " astraeiform " colonies are produced, which do not difFer essentially from those formed by budding. Frequently, however, individuals formed by fission become only imperfectly separated, remaining proximally more or less closely confluent. In such cases the calices form continuous, straight, curved or labyrinthic furrows, with more or less clearly distinguishable centres. The Compound corallum of a polyp stock remains practically the same as in solitary individuals, excepting that the conditions are more complicated when the Separation of the zooids is incomplete. Dendroid and massive colonies frequently develop a common connective matrix or tissue (coenenchyma) which unites the various corallites into a whole; it is secreted by the common colonial flesh, called coenosarc, which extends as a carpet between the polyps, The coenenchyma is sometimes dense in structure {Oculinidae), or it may consist of a vesicular or tubulär tissue. The separate corallites are often also united by means of the septa, which are produced over and beyond the thecae, and fused with those of neighbouring individuals. In such cases the interseptal loculi are almost always filled with strongly developed dissepiments. All structures developed in the included space within the theca, with the exception of the septa and columella, are designated coUectively as endotheca ; those lying without the theca as exotheca. The Anthozoa are exclusively marine forms, and predominate in shallow water. Many of the Adiniaria, Antipatharia and Madreporaria occur also at greater depths, ranging from 50 to 300 and sometimes to over 3000 fathoms. The so-called reef-corals inhabit depths usually not exceeding 45 metres, and require a temperature of the water of 20° C, or higher. Hence, existing coral-reefs are restricted to a zone exfcending about 30" on either side of the equator ; they are distinguished according to form as fringing reefs, barrier reefs and atolls. While the stony corals (Porifes, Acropora, Turbinaria, Pocillopora, numerous '^ Astraeidae" and Fungidae) and the Alcyonarians (Heliopora) are the most important, they are not the only agents concerned in the formation of reefs, as an active part is also played by the Hydromedusae (Milleporidae), calcareous algae (Lithothamnium, Melobesia), mollusks, echino- derms, bryozoans and worms. Of the ancient coral-reefs which have been formed in nearly all of the great geological periods, those of the Cenozoic and Mesozoic periods are composed in part of genera similar to those now living ; while those of the Paleozoic represent genera and families that are now principally extinct, and whose relation to living forms is often quite uncertain. The Anthozoa are divided by Haeckel into three subclasses : Tetracoralla, Hexacoralla and Alcyonaria or Octocoralla. Of these the two first-named groups are by some authors collectively termed Zoantharia. 80 COELENTERATA— ANTHOZOA phylüm ii Subclass 1. TETRAOORALLA Haeckel.^ {Zoantharia Eugosa Milne Edwards ; Fterocorallia Frech.) Extinct, Faleozoic, simple or composite sderodermic corals, with septa arranged according to a tetrameral System, and either Ulaterally or radially symmetrical ; wühout coenenchyma, hut with usually strongly developed endothecal tissue in the form of tabulae or dissepiments, and with well-marked, frequently wrinkled epithecal wall. The Tetracoralla are especially characterised by having the septa, subse- quent to the formation of the primaries, introduced along four lines rising from the apex of the base of the corallum. The earlier stages of the Tetra- coralla have recently been reinvestigated by Duerden, Carruthers and others. Duerden concluded that the observations of Ludwig and Pourtales on the primary hexamerism of these corals were correct. According to Carruthers, in the developing young Tetracoralla the first stage of septal formation is for a Single septum to stretch entirely across the calice from wall to wall. This septum, which is called the axial septum, later breaks up to form the main (cardinal) and the counter septum of the mature coral. In the next stage a small septum appears on each side of the main septal end of the axial septum. These two septa form the alar septa of the mature corallum. In the third stage two other septa appear, one on each side of the counter septal end of the axial System. After the formation of these six septa there is a distinct pause in the formation of new septa and any irregularity in the disposition of the septa is corrected. Four of the six septa are called principal, and are con- spicuous in the later septal arrangement ; these four are the main, counter, and alar septa. Two of the first six septa, one on each side of the end of the counter septum, are not so prominent in subsequent development. There is a controversy as to whether the primary septa of the Tetracoralla are four or six in number : Duerden and Carruthers holding the number to be six, while Brown and Gordon contend that it is four. The four principal septa are sometimes of equal proportions, when they may be either stouter and longer than the others (Stauria), or thinner and shorter (Omphyma) ; or they may be of unequal proportions. Of the two principal septa which lie in the longitudinal axis of the corallum, one (called the main or cardinal septum) is frequently situated in a depression or furrow known as the fossula (Fig. 100); while the other or counter septum is either normally developed, or is more or less reduced. Occasionally the counter septum is placed in a fossula, while the cardinal septum is normally developed ; but the two laterally disposed or alar septa are always equal in size. The remaining septa not infrequently exhibit a well-marked radial arrangement, in which the longer and more strongly developed usually alternate with the shorter and less strongly developed. New septa, according to Kunth and Dybowski, are inserted in the foUowing order. First, a new septum is given ofF on either side of the 1 Literature : Kunth, A., Beiträge zur Kenntniss fossiler Korallen. Zeitschr, deutsch, geol. Ges., 1869-70, vols. xxi., xxii. — Dybowski, W. N., Monographie der Zoantharia Eugosa, etc. Archiv für Naturkunde Liv-, Est-, und Kurlands, 1874, vol. v. — Roemer, F., Lethaea Palaeozoica, 1883, pp. 324-416. — Schlüter, dem., Anthozoen des rheinischen Mittel-Devons. Abhandl. j^reuss. geol. Landes-Anstalt, 1889, vol. viii. — Brown, T., Studies on the Morphology and Development of certain Rugose Corals. Ann. N.Y. Acad. Sei., 1909, vol, xix. — Faurot, L., Affinites des Tetracoralliaires et des Hexacoralliaires. Annales de Paleont., 1909, vol. iv. i... TETRACORALLA 81 Cardinal septum (Fig. 100, Ä), and takes up a position parallel with the alar septum. This leaves an intermediate space between the cardinal and the newly formed septa, which becomes filled, however, by the repeated insertion of new septa one above the other in the same manner as the first ; and hence they diverge from the cardinal septum, as they grow upward, in a pinnate fashion. Likewise the two counter quadrants lying between the alar and counter septa become occupied by lamellae which are given ofF from the alar septa, and gradually arrange theraselves parallel with the counter septum. The mode of growth in the Tetracoralla will be readily understood on inspect- ing the surface of those specimens, the septa of which are visible on« the exterior, or where the wall is readily removed by corrosion or polishing. One may then note three distinct lines extending from the calicinal margin to the base ; these mark the cardinal and the two alar septa, from which the other pinnately branching septa are directed obliquely upward (Fig. 101). The Order in which the septa are given ofF in the four quadrants, according to Kunth, is indicated by the numerals in Fig. 100. Many of the Tetracoralla multiply only by sexual reproduction, and occur only as single individuals ; asexual reproduction takes place usually by calicinal, more rarely by lateral gem- mation, and results in dendroid or massive colonies. ' Dissepiments are generally abund- antly developed between the septa, which latter are compact, and the Upper edges of which are either smooth or serrated. Sometimes the dissepiments fiU the whole interior with a vesicular tissue, and the central visceral cavity is frequently entirely partitioned off by horizontal, inclined or funnel-shaped tabulae. The wall is usually composed of the thickened and fused septal edges; sometimes it is invested with epitheca and furnished with vertical rugae or root-like processes. A true coenenchyma is absent. In a few genera the calice is provided with a lid or operculum, which may be composed of one (Calceola) or of several plates (Goniophyllum), With the exception of a few genera the systematic position of which is uncertain, all the typical Tetracoralla are confined to the Paleozoic rocks. Family 1. Oyathaxonidae Milne Edwards and Haime. Turbinate or horn-shaped simple coralla. Septa with regulär radial arrange- ment. Tabulae and dissepiments absent. Silurian to Permian. Cyathaxonia Mich.. (Fig. 102). Acutely pointed, conical. Cardinal septum in fossula. Septa numerous, extending inward as far as the strongly developed styliform and considerably elevated columella. Carboniferous limestone ; Belgium and England. VOL. I 0 Fig. 100. Menophyllum tenuimar- ginatum B. and H. Car- boniferous Limestone ; Tournay, Belgium. 2/j. h, Cardinal septum ; g, Counter septum ; s, Alar septa. Streptelasma profundum (Owen). Ordovician ; Cin- cinnati, Ohio. Natural size. 82 COELENTEEATA— ANTHOZOA PHYLUM II Duncanella Nich. Corallum top-shaped. Septa nearly all of uniform length and size, forming a spurious columella in centre of the deep calice, exsert at the base. Silurian ; North America. D, horealis Nich. Fetraia Münst. (Fig. 103). Turbinate or conical. Septa short, reaching to the centre only at the base of the very deep calice. Columella absent. Ordovician to Carboniferous. Folycoelia King (Fig. 104). Horn-shaped. Calice very deep; four prin- cipal septa reach nearly to its centre, between which in each quadrant are five shorter septa. Zechstein. Kanophyllum Dyb. Ordovician and Silurian. Fig. 102. Cyathaxonia cornuMich. Car- boniferous Lime- stone ; Tournay, Belgiiim. Cor- allum with fractured theca, showing open in- terseptal loculi. 2/1. Fig. 103. Petraia radiata Münster. Devonian; Enkeberg, near Brilon, i/i. a, Cor- allum viewed from the apex; b, Trans- verse section below the middle. Fig. 104. Folycoelia pro- funda (Germ.). Zech stein; Gera. Vi(after Roemer). Fig. 105. Falaeocydus porpita (Linn.). Silurian; Gotland. a, Top View of calice ; h, Profile, i/i- Fig. 106. Microcyclus dis- cus Meek and Worth. Hamil- ton (Devonian) ; North America. i/j. a, Corallum from below ; b, from . above (after Nicholson). . Family 2. Palaeocyclidae Dybowski. Coralla simple, discoidal or howl-shaped. Septa numerous, stout, approaching radial symmetry in disposition. Tabulae and dissepiments waiiting. Palaeocyclus E. and H. (Fig. 105). Discoidal to depressed top-shaped, with epitheca. Septa numerous, radially disposed, the larger ones reaching to the centre. Silurian. Type, P. porpita (Linn.). Combophyllum, Baryphyllum E. and H. Devonian. Hadrophyllum E. and H. Cushion-shaped, with epitheca. Calice with three septal fossula, that of the cardinal septum being the largest, Devonian; Eifel and North America. Microcyclus Meek and Worth. (Fig. 106). Like the preceding, but with only one septal fossula. Devonian ; North America. Family 3. Zaphrentidae Milne Edwards and Haime. turbinate, conical or cylindrical ; septa numerous, exhihiting Theca generally formed hy fusion of s not very ahundant in inter- Coralla simple, distinct bilateral symmetry in arrangement. septal ends. Tabulae completely se Streptelasma Hall (Fig. 107). Turbinate, often curved. Septa numerous rBCLASS I TETRACORALLA 83 10-130), alternately long and short; the free edges of the longer septa are 'isted together in the centre to form a pseudo-columella. Tabulae few er )sent. Position of the cardinal septum is recognisable on the exterior by the System of pinnately diverging costal ridges. Common in Ordovician and Silurian. S. pro- s a k fundmn (Owen), the type species, has often been confused with S. corniculum and various species of Zaphrentis. Zaphrentis Raf . (Caninia Mich. pars) (Figs. 108- 10). Simple, tur- binate or sub- cylindrical, fre- quently elongated. Calice deep, with circular margin. Septa numerous, reaching to the centre ; cardinal septum in a deep fossula. Tabulae numerous; somewhat irregulär, and passing from side to side of the visceral chamber ; dissepiments sparingly developed in outer zone of corallum. 50 to 60 species known, ranging from Silurian to Carboniferous. Maximum development in Carboniferous. Amplexus Sow. Simple, sub-cylindrical or elongated turbinate. Calice Streptelasma profundum (Owen). Cincinnatian Group (Ordovician); Cincinnati, Ohio. Side view. B, Transverse section, C, Longitudinal section. (h, Cardinal septum ; g, counter septum ; s, alar septum.) Fig. 108. Zaphrentis cornlcxda Lesueur. Devonian lime- stone ; Ohio. Fig. 109. Zaphrentis cornucopiae Mich. Calice enlarged. Carboniferous Limestone ; Tournay, Belgium. Fig. 110. Zaphrentis enniskilleni Nich. Carboniferous Lime- stone ; A, B, Transverse sections through respectively Upper and lower portions of calice. (', A long and two Short septa united at the ends to form the wall. D, Longitudinal section showing tabulae (after Nicholson). shallow, usually with septal fossulae. Septa moderately numerous, short, never produced to centre. Tabulae highly developed, horizontal. Ordovician to Lower Carboniferous. Type, A. coralloides Sow. Aulacophyllum E. and H. Turbinate. Septa numerous, extending to 84 COELENTERATA— ANTHOZOA PHYLUM II centre. Cardinal septum in deep fossula ; adjacent septa pinnately developed. Ordovician to Devonian. Menophyllum E. and H. (Fig. 100). Turbinate. Cardinal septum in largest of three fossulae. Lower Carboniferous limestone. Lophophyllum E. and H. Carboniferous limestone. Anisophyllum E. and H. Ordovician to Devonian. Pycnophyllum Lindstr. Ordovician and Silurian. Apasmophyllum Eoem. Metriophyllum E. and H. Thamnophyllum Penecke. Devonian. Penfaphyllum de Koninck. Carboniferous. Family 4. Cyathophyllidae Milne Edwards and Haime. Simple or composite coralla. Septa numerous, radially arranged ; the four piincipal septa rarely distinguished by greater or smaller size. Tabulae and vesicular tissue (dissepiments) abundant. Cyathophyllum Goldf. (Figs. 111-13). Extremely variable in form, sometimes simple, turbinate or sub-cylindrical ; sometimes giving rise to bushy, fasciculate or astraeiform colonies, where reproduction takes place by Fig. 111. Cyathophyllum caespitosum Goldf. Devonian ; Gerolstein, Eifel. Natural size. Fig. 112. Cyathophyllum hexagonum Goldf. Devonian stein, Eifel. Natural size. Gerol- calicinal or lateral gemmation. Septa very numerous, strictly radial in arrangement, and of ten alternately long and short ; the longer septa extend- ing to the centre. Visceral Chamber filled with numerous imperfectly developed tabulae ; vesicular dissepiments highly developed in peripheral portion. Nearly 100 species known, ranging from Ordovician to the Lower Carboniferous. Maximxum development in Devonian. Campophyllum E. and H. (Fig. 1 1 4). Like the preceding, but septa not extending to the centre. Devonian and Carboniferous Limestone. Heliophyllum Hall. Usually simple and turbinate, more rarely forming dendroid colonies. Septa numerous, extending to the centre, and thickened on their sides by conspicuous vertical ridges (" carinae "). Devonian. Diphyphyllum Lonsd. (Fig. 115). Ordovician to Carboniferous. Pholido- lum Lindstr. Ordovician and Silurian; Eridophyllum E. and H. Silurian FBCLASS I TETEACORALLA 85 I ^^pd Devoniaii. Crepidophyllum Nich. Craspedophyllum Dybowski. Devonian. ^Ktoninckophyllum Nich. Chonaxis E, and H. Carboniferous. Clisiophylhm ^^Pana. Silurian to Carboniferous. ^H Oinphyma llsiL (Fig. 116). Corallum simple, conical or turbinate ; theca I Fig. 113. Cyathophyllum heteropliyllum E. and H. Middle De- vonian ; Gerolstein, Eifel. A, Transverse ; B, Longi- tudinal section (after Nicholson). Fig. 114. , Campophyllum com- pressum Ludw. Car- boniferous Limestone ; Hausdorf, Silesia. a, Longitudinal ; b, Trans- verse section. Fig. 115. Diphyphyllum con- cinnum Lonsd. Car- boniferous Lime- stone ; Kamensk, Ural. with root-like processes. Septa numerous ; the four principal septa in shallow fossulae. Surface marked with pinnafcely branching Striae. Tabulae numerous. Silurian. Chonophyllum E. and H. Silurian and Devonian. Fig. 116. Oinphyma subturbinata B. and H. Silurian limestone Gotland, Svveden. a, Side view ; h, Calice from above. Fig. 117. Lithostrotion martini E. and H. Lower Carbonifer- ous ; Hausdorf, Silesia. Sections of individual corallite (after Kunth). Ptychophyllum E. and H. Simple and turbinate, or composite. Each stock is composed of funnel-shaped, invaginated layers, representing calicinal buds, the marginal lips of which are more or less reflected outwards. Septa numerous and strongly twisted in the centre to form a pseudo-columella ; their 86 COELENTERATA— ANTHOZOA PHYLUM II peripheral edges are thickened and are fused with one another so as to form a wall. Silurian (P. patellatum Schlot, sp.) and Devonian. Cyclophyllum Duncan and Thom. Simple, cylindro-conical. Septa numer- ous, the longer ones forming a thick pseudo-columella with enclosed spongy tissue. Aulophyllum E. and H. ; Asjpidophyllum, Bho- dophyllum Nich. and Thoms., etc. Carboniferous. Lithostrotion Llwyd (Stylaxis M'Coy ; Fetalaxis E. and H.) (Fig. 117). Fasciculate or astraei- form Stocks composed of prismatic or cylindrical corallites. Septa numerous, alternately long and short. Styliform columella in the centre. Abundant in Carboniferous limestone. Fig. 118. Lonsdaleia floriformis Lonsd. Car- boniferons Limestone ; Kildare, Ire- land. i/i. a, Two cylindrical corallites, partially splitopen ; h, Twohexagonal calices, seen from above. Fig. 119. Phillipsastrea hennaM (Lonsdale). Devonian limestone ; Ebersdorf, Silesia. a, Upper surface ; h, Transverse section. Natural size. Lonsdaleia M'Coy (Fig. 118). Fasciculate or astraeiform, composite coralla. Septa well developed ; columella large, composed of vertically rolled lamellae. Central tabulate area bounded by an interior dissepimental wall, between which and the theca vesicular endotheca is abundantly developed. Common in Carboniferous rocks. Strombodes Schweigg. Astraeiform Stocks composed of small prismatic corallites. Septa extremely numerous, very slender, extending to the centre. Fig. 120. Stauria astraeiformis E. and H. Silurian ; Gotland, Sweden. A, Transverse section parallel to upper aurface. B, Enlarged transversa section of individual corallite. C, Several calices from above. Natural size (after Nicholson). Theca imperfectly developed. Visceral Chamber filled with infundibuliform tabulae and vesicular tissue. Silurian (S. typus M'Coy sp.) and Devonian. Fachyphyllum, Spongophyllum E. and H. Silurian and Devonian. Acervularia Schweigg. Astraeiform or bushy colonies. Septa stout and SUBCLASS I TETRACORALLA 87 (lumerous. An iiiterior wall is present ; tabulae are developed in the central ;irea, while the peripheral zone is filled with vesicular tissue, Silurian (A. (Dianas Linn. sp.) and Devonian. PhiUipsastrea d'Orbigny (Fig. 119). Astraeiform colonies, with indi- vidual corallites united by confluent septa, which are produced beyond the theca, and obscure the same. Interseptal loculi filled with vesicular endotheca. Devonian and Carboniferous. Type, P. hennahi (Lonsd.). Stauria E. and H. (Fig. 120). Astraeiform or bushy composite coralla. Septa well developed; the four principal septa characterised by larger size, and forming a complete cross in the centre of each corallite. Silurian (Wenlock). Columnaria Goldf. (Favistella Hall). Astraeiform Stocks, composed of long, polygonal, thick-walled corallites. Septa radially arranged in two cycles, alternately long and short, barely reaching the centre. Tabulae horizontal, disposed at regulär intervals apart, and stretching across the entire visceral Chamber. Dissepiments imperfectly developed or absent. Ordovician to Devonian. Ileterophyllia M'Coy. Carboniferous. Batiersbyia E. and H. Devonian. Family 5. Oystiphyllidae Milne Edwards and Hainie. Usually simple coralla. Septa very thin ; interseptal loculi filled with vesicular endotheca or compact stereoplasma. Tabulae absent ; central area of visceral Chamber either completely filled with vesi- cular tissue or stereoplasma, or containing the same only in the lower portions of Chamber. Cal- careous operculum sometim,es present. Cystiphyllum Lonsd. (Figs. 121, 122). Simple, very rarely forming bushy colonies. Calice deep; the entire visceral Cham- ber filled with vesicular tissue, which, as a rule, wholly obliterates the numerous radi- ally directed septa. Silurian and Devonian. Strephodes M'Coy (Fig. 123). Usually simple coralla. Septa well developed, alter- nately long and short, some- times forming a pseudo- columella. Silurian to Carbo- niferous. Goniophyllum E. and H. (Fig. 124). Corallum simple, in the form of a four-sided pyramid, and covered with thick epithecal tissue. Calice deep ; septa numerous, thick and very short. Entire visceral Chamber filled with vesicular and stereo- plasmic endotheca. Operculum composed of four plates symmetrically paired. Silurian. Fig. 121. Fig. 122. Cystiphyllum cylindricum Lonsd. Silurian ; Iron Bridge, Cystiphyllum vesiculosum England. A, B, Transverse Goldf. Devonian ; Eifel. and longitudinal sections (after Natural size. Nicholson). 88 COELENTERATA—ANTHOZOA PHYLUM II Bhizophyllum Lindst. Corallum simple, pyramidal or hemispherical, flattened on one side ; external surface corrugated, and sending ofF hollow, root-like epithecal processes. Calice marked with septal Striae ; internal structure consisting of vesicular tissue and stereoplasma. Operculum in form of semicircular plate ; inner surface traversed by median ridge and fainter, granulated, parallel elevations. Silurian. Calceola Lam. (Fig. 125). Corallum simple, semi-turbinate, or slipper-shaped, Fio. 123. Strephodes mnrchisoni Lonsd. Showing strongly developed dissepiments and tabulae. Fk.. 124. Goniophyllum pyramidale (His.). Silurian ; Gotland. A , Specimen with operculum. B, Calice seen from above. Natural size (after Lind- ström). Calceola Devonian ; Fig. 125. sandalina Lam. Eifel. Natural with one side flat and triangulär. Calice very deep, extending nearly to apex, and marked internally with fine septal Striae. Cardinal septum placed in the centre of the vaulted side, counter septum in middle of flattened side, and alar septa at the angles. Internal structure composed of fine vesicular tissue and Stereoplasma. Operculum semicircular, very thick, under surface marked with prominent median and fainter lateral septal ridges. C. sandalina Lam. Very common in Middle Devonian of Europe, rare in Carboniferous Limestone of Belgium. Range and Distribution of the Tetracoralla. The typical Tetracoralla are confined to the Paleozoic rocks. They are unknown in the Cambrian, and make their first appearance in the Ordovician, where they are sparsely represented in North America and in Europe. Here the most abundant genus is Streptelasma, and next in order of importance are Cyathophyllum, Ftychophyllum and Columnaria. The maximum development falls in the Silurian, which contains the largest number of genera and species. There are limestones found on the Islands of Gotland and Dago (Esthonia), as well as at Dudley, Shropshire, at Lockport, New York and other places in North America, which are made up of ancient coral-reef s. The principal agents concerned in the formation of these reefs were Cyathophyllum, Heliophyllum, Omphyma, Ptychophyllum, Sfromhodes, Acervularia, Stauria, Aulacophyllum, Cysti- phyllum, etc., of the Tetracoralla, besides numerous Tabulata, Octocoralla, süBCLÄSS II HEXACORALLA 89 Bryozoa, and Echinoderms. The Tetracoralla are not less conspicuous in the Devonian, especially in the Middle and Upper Devonian of the Eifel district, Westphalia, Nassau, Harz, Boulogne, England, and North America. ■- Particu- larly abundant here are the genera Cyathophyllum, Campophyllum, Zaphrentis, Cystiphyllum, Phülipsasti'ea, Calceola, etc. Zaphrentis, Amplexus, Lithostrotion, Lonsdaleia, Cyclophyllum^ etc., predominate in the Carboniferous Limestone of Belgium, England, Ireland, and North America ; while in the Zechstein the solitary genus known is Polycoelia. On the other band, the Permo-Carboni- ferous rocks of the Salt Range in India and of the island of Timor contain the genera Zaphrentis, Amplexus, Clisiophyllum, and Lonsdaleia. According to Frech, the genera Gigantostylis, Piriacophyllum, and Coccophyllum, occurring in the Alpine Trias, belong to the Tetracoralla ; and to this group also have been assigned Holocystis E. and H., from the Cretaceous, and the recent genera Haplophyllum Pourtales, and Guynia Duncan. A number of Paleozoic Tetracoralla, such as Baitersbyia, Heterophyllia, and Stauria, are referred by Duncan and Nicholson to the Hexacoralla (^' Astraeidae"). Subclass 2. HEXACORALLA Haeckel. (Zoantharia Blainville ; Hexadinia and Polyadinia Ehrenberg.) Simple or composite polyps, with radial mesenferies arising in cydes of six, iwelve, or multiples of six (more rarely pentameral, septameral or odameral) ; frequently with calcareous corallum, hut sometimes fleshy or with hoo'ny axis. To the Hexacoralla belong the calcareous reef-building and deep-sea corals (Madreporaria) of the present day, the fleshy sea-anemones (Adiniaria), and those forms characterised by the secretion of a horny axis (Antipatharia). Of these three Orders, only the Madreporaria are known in a fossil State. These forms are distinguished from the Tetracoralla by the hexameral System and radial arrangement of mesenteries and septa ; and from the Octocoralla, in addition to the above-named characters, by their simple tentacles. According to Duerden, either before or shortly after extrusion of the larva, the six primary pairs of mesenteries (protocnemes), constituting the first cycle, make their appearance. The organs arise in bilateral pairs, in a regulär and well-defined order, which is uniform for all the species yet studied. The first two or three pairs arise around the oral extremity of the larva, while the others first appear at varying distances down the wall. The protocnemic sequence is represented by the Roman numerals in Fig. 126, and agrees with that established for the greater number of actinians. The first four pairs very early unite with the stomodaeum, but the fifth and sixth pairs remain free or incomplete for a lengthened period, suggesting a diflferent phylogenetic significance from the others. The six pairs of second cycle mesenteries (metacnemes) arise after fixation, but in a manner altogether difFerent from that followed by the first cycle. They appear on the polypal wall in unilateral pairs or couples within the six primary exocoeles, and in a succession which is from the dorsal to the ventral side of the polyp, not the whole cycle at a time. For a long time, as shown in Fig. 127, the six pairs present a diff"erence in size, corresponding with their dorso-ventral or antero-posterior order of appearance. 90 COELENTERATA— ANTHOZOA PHYLÜM II The twelve pairs of third cycle mesenteiies are found to develop in a succession fwhich is altogether unexpected. They follow the same dorso- ventral order as the second cycle pairs, but in two series. A primary series of six pairs — one pair within each sextant — appears within the exocoele on the dorsal aspect of each of the second cycle mesenteries, one pair following upon another, and then another series of six pairs arises on the ventral aspect of the second cycle mesenteries in the same order (Fig. 128). In the later stages of development the regularity of the mesenterial succession is not IZ IE w JS. m m Fig. 126. Growth stages of coral polyp in Maeandra (" Manicina") areolata. Diagrammatic flgures showing order of appearance of the six primary pairs of mesenteries. In a only two pairs of mesenteries are present, of which one pair (i) is imited with the stomodaeum, while the other (ii) is free ; in b the second pair of mesenteries has become complete, and a third pair (iii) has appeared on tlie ventral border ; in c another pair (iv) is found within the dorsal Chamber ; in d the first four pairs of mesenteries to arise have all become complete, and the fifth and sixth pairs (v, vi) have appeared, but remain incomplete for a long period, the secondary mesenteries appearing in the meantime (cf. Fig. 127). The actual stages given are taken from Maeandra areolata, but a like sequence is presented by other species whose development has been foUowed (after Duerden). always maintained ; one region may be somewhat in advance of, or may lag behind its normal development. The sequence thus outlined in the briefest manner is sufficient to show that the development of the mesenteries in coral polyps is bilateral, and takes place in stages from one extremity to the other. The radial symmetry, characteristic of the adult polyp, is thus derived from primitively bilateral Organs, which appear in an antero-posterior succession. Moreover, each cycle represents a separate period of development, as compared with the successive growth in one direction of ordinary segmented animals. The first two cycles of tentacles (prototentacles) generally arise a cycle at a time, either simultaneously or one following the other. The later tentacles SlinCLASS II HEXACORALLA 91 are developed in an order in correlation with that of the mesenteries, some- times entocoelic and exocoelic members appearing together. In the process of growth the exocoelic members are always relegated to the outermost cycles, in a manner first established by Lacaze-Duthiers for actinians ; only the entocoelic tentacles are of any ordinary value, Siderastrea radians (Pallas) is exceptional in that the exocoelic tentacles appear in advance of the entocoelic. Fig. 127. Growth stages of larval polyps in Siderastrea radians. Three diagramrnatic figures illustrating the manner of appearance of the six mesenteries (A-C) constituting the second cycle. The mesenteries arise in unilateral pairs witliiii corresponding exocoelie Chambers on each side of the polyp. At first (a) a ])air appears within the dorso-lateral exocoele on each side ; shortly after (li) a similar pair arises within each middle exocoele ; then (c) a pair within each ventro-lateral exocoele. For a long time the pairs retain a diö'erence in size, corresponding with their order of appearance (after Duerden). The skeleton never appears until after fixation of the larva. It makes its first appearance in the form of minute plates or granules, as an ectoplastic pro- duct of the ectodermal cells (calicoblasts) of the base. A flat, circular, basal plate is formed by the union of these, and may later become produced upward at the edge as the epitheca, while from its inner or polypal surface the septa begin to appear as vertical upgrowths formed within invaginations of the basal disk of the polyp. The skeletal cup first formed is known as the prototheca. 92 COEliENTERATA— ANTHOZOA PHYLUM II Like the tentacles, the first two cycles of septa (protosepta) may appear simultaneously, or the cycle of six entosepta may arise in advance of the cycle of six exosepta. The order of appearance of the later cycles is not yet thoroughly understood, the relative sizes in the mature corallum by no means indicating the actual order of developnient. As in the case of the mesenteries, the radial plan of the mature septa is derived from structures which appear bilaterally, in a more or less definite dorso - ventral or antero - posterior Fig. 128. Three stages in the development of the twelve pairs of third-cycle mesenteries. All the six pairs of primary mesenteries are now complete, and the second-cycle pairs are all equal, but free from the stoniodaevnn. In a a pair of third-cycle mesenteries (iii) has appeared on each side, within the exocoele next the dorsal directives ; in b a corresponding pair occurs within the dorsal of the two exocoeles of all the six Systems, the order being from the dorsal to the ventral aspect ; in c another series of six pairs is beginning, situated within the ventral of the two exocoeles in each system. Growth in the dorsai region is in advance of that in the ventral (after Duerden). succession. Furthermore, as in the case of the tentacles, the exosepta remain exosepta throughout the course of their development, always constituting the outermost cycle. The entosepta beyond the primary six follow the same succession of growth as the mesenteries, so that the order assigned the secondary and tertiary mesenteries in Fig. 129 will also hold for the septa. Eeproduction takes place either sexually, when separate individuals are pro- duced ; or asexually, by means of lateral or basal gemmation ; or by fission. In composite coralla, the individual corallites are sometimes united by a suBCLASs II HEXACORALLA 93 common coenenchyma. Endothecal structures are frequently present in the form of synapticulae, dissepiraents, and tabulae. The Order of stone corals or Madreporaria {Zoantharia sclerodermata) was divided by Milne Edwards and Haime into five suborders : Eugosa, Tahulata, Tuhulosa, Perforata, and Aporosa. Of these, the Eugosa have been elevated by Haeckel into a separate subclass under the name of Tetracoralla. The groups Aporosa and Perforata are called Hexacoralla ; while the afiinities of the Fig. 129. Diagram showing the Order of appearance of all the mesenteries in a polyp having three cycles. The Roman numerals represent the cycles to which the mesenteries belong, and the smaller Arabic numerals indicate the Order in which the mesentery appeared within its cycle. The regularity here indicated Is constant for the primary and secondary cycles, but departure may be encountered in the third cycle (after Duerden). Tahulata (with which the Tuhulosa are now generally included) are still unsatisfactorily determined. The group is certainly composed of a varied assemblage of forms, some of which have been assigned to the Hexacoralla, some to the Octocoralla, and some to the Hydrozoa and Bryozoa. Order 1. MADREPORARIA Milne Edwards. ^ (Zoantharia sclerodermata E. and H.) Eadially symmetrical sclerodermous corals with typically hexameral (rarely pentameral, heptameral, or octameral) arrangement of septa. ^ Literature : Pratz, E. , Ueber die verwandtschaftliche Beziehungen einiger Korallengattungeu, etc. Palaeontogr. 1882, vol. xxix. — Frech, F., Die Korallenfauna der Nordalpinen Trias. Palaeontogr. 1890, vol. xxxvii. — British Museum Cat. of Madreporarian Corals, vol. i. by George Brook, 1893, vols. ii.-vi. by //. M. Bernard, 1896-1906.— Volz, W., Die Korallen der Schichten von St. Cassian in Süd-Tirol. Palaeontogr. 1896, vol. xliii. Felix, J., Anthozoen der Gosauschichten in den Ostalpen. Palaeontogr. 1903, vol. xlix. — Duerden, J. E., The Coral Siderastrea, etc. Carnegie Inst. Wash., 1903, Pub. No. 20.— Lang, W. D., Growth-Stages in the Coral Genus Para- smilia. Proc. Zool. Soc. London, 1909, pt. ii. — Paroim, G. F., La Fauna coralligena del Cretaceo dei Monti d' Ocre nell' Abruzzo. Mem. Com. Geol. Ital., 1909, vol. v. (See also ante, p. 74). 94 COELENTERATA— ANTHOZOA PHYLÜM II Suborder 1. APOROSA Milne Edwards and Haime. Septa and theca compact; interseptal loculi usually partitioned off hy dissepi- ments or synapticulae, more rarely hy tabulae, seldom empty throughout. Theca either independenÜy secreted, or formed hy fusion of the septal or ahsent. Family 1. Turbinolidae Milne Edwards and Haime. Corallum simple, very seldom composite ; septa numerous, long, and with entire margins. Interseptal loculi empty throughout. Columella usually, pali often present. Theca comi The Turhinolidae begin in the Jurassic, and are especially abundant in the Tertiary and at the present day. Sexual reproduction prevails, although a few forms multiply by gemmation ; the buds, how- ever, become separated from the parent animal at an early period. Turhinolia Lam. (Fig. 130). Corallum free, conical, with circular calice. Septa produced beyond the theca. Styliform Fig. 130. Turhinolia bowei-- lanJci E. and H. Eocene ; Highgate, England. 6/^. Fig. 131. Ceratotroclius duodecimocostatus (Goldf.). Miocene Baden, near Vienna. Natural size. Fig. 132. Fläbelhim roissyanum E. and H. Miocene; Baden, near Vienna. Natural size. columella present. Tertiary and Recent ; common in Calcaire Grossier of the Paris Basin, and Eocene of England and southern United States. Sphenotrochus E. and H. Free, cuneiform with elongated calice; colum- ella lamellar. Cretaceous to Recent. Type, S. crispus (Lam.). Eocene to Recent. Common in Calcaire Grossier of the Paris Basin, and in the Eocene of the Gulf States. Smilotrochus E. and H.; Stylotrochus From.; Onchotrochus Duncan. Cre- taceous. Discotrochus E. and H. etc. Tertiary. Ceratotrochus E. and H. (Fig. 131). Horn-shaped ; young forms attached at the apex. Septa very numerous, produced above the theca; columella fasciculate. Cretaceous to Recent. Flahellum Lesson (Fig. 132). Corallum wedge-shaped, compressed, free, or attached. Septa numerous. Wall covered with epitheca, and sometimes furnished with spinous processes. Tertiary and Recent. Trochocyathus E. and H. (Fig. 133). Horn-shaped, with circular calice. Septa stout ; columella papillous and trabecular, and surrounded by several cycles of pali. Numerous species from Lias to Recent. ;übclass II HEXACORALLA 95 Thecocyathus E. and H. Depressed, conical, or discoidal, attached early in life, later becoming free. Wall with thick epithecal Investment. Calice circular, septa numerous ; columella fasciculate, and sur- rounded by several cycles of pali. Lias, Jurassic, Cre- taceous, and Recent. Paracyathus, Deltocyathus E. and H. (Fig. 134). Tertiary and Recent. Discocyathus E. and H. Jurassic. Coenocyathus, Acanthocyathus, JBathycyathus E. and H., etc. Tertiary and Recent. Caryophyllia Lam. (Fig. 135). Turbinate, with broad Fig. 133. Fig. 134. Trochocyathiis conulus From. Aptian ; Deltocyathus italicus E. and H. Miocene ; Fio. 135. Caryophyllia cyathus Sol. Recent. Lougitudi- Haute Marne, a, Profile, natural size ; h, Porzteich, Moravia. Calice enlarged. size ; b, Calice enlarged. Profile, natural nal section, natural size (after Milne Edwards). base, attached. Calice, circular ; columella papillous, trabecular, and sur- rounded by a single cycle of pali. Cretaceous to Recent. Family 2. Oculinidae Milne Edwards and Haime. Invariably composite coralla, increasing by lateral gemmation. Walls of corallites thickened by a compact coenenchyma. Lower portion of visceral Chamber narrowed or filled up by deposition of stereoplasma. Septa moderately numerous ; interseptal loculi usually open to the base. Lias to Recent ; fossil forms not particularly numerous, Oculina Lam. Corallites irregularly or spirally distributed over the smooth surface of coenenchyma. Septa slightly projecting ; columella papillous, surrounded by cycle of pali. Tertiary and Recent. Agathelia Reuss. Like the preceding, but form- ing tuberous or lobate colonies. Cretaceous and Tertiary. Synhelia E. and H. Cretaceous. Astrohelia E. and H. Tertiary. Psammohelia, Euhelia E. and H., etc. Jurassic. Ilaplohelia Reuss. Small, arborescent, with corallites all disposed on one side of the branches. Coenenchyma striated or granulated. Septa in three cycles ; columella and pali present. Oligocene. Enallhelia E. and H. (Fig. 136). Stock branch- ing ; corallites disposed usually in alternating sequence in two rows along the sides of branches. Coenenchyma highly developed, striated, or granulated; columella rudi- mentary. Jurassic. Type, E, Compressa (d'Orb.). Fig. 136. Enallhelia striata Quenst. Coral- Rag ; Nattheim. a, Natural size ; b, Calice enlarged. 96 COELENTERATA— ANTHOZOA PHYLUM ir Family 3. Pocilloporidae Verrill. Composite, branchirig, lohate, or massive colonies, with small cylindrical corallites, united hy compact coenenchyma. Sepfafew (6-24), sometimes rudimentary. Visceral Chamber partitioned off by horizontal tabulae. Of the two Recerit genera belonging to this family, Pocillopora and Seriato- pora Lam., the former occurs also in the Miocene of the West Indies. Family 4. Stylophoridae Milne Edwards and Haime. Composüe coralla, with corallites united by vesicular or compact coenenchyma. f^ 1, Septa well developed ; a prominent, styli- form central columella; interseptal loculi empty throughout. Jurassic to Eecent. Stylophora Schweigg. (Fig. 137). Stock branching, or depressed, massive and tuberous. Calices small. Fig. 137. Stylophora subreticulata Reuss. Miocene ; Grund, near Vienna. a, Corallum, natural size; b, surface greatly enlarged. Araeacis E. and H. Eocene. embedded in spinous, coenenchyma. developed, moderately columella styliform. Tertiary, and Recent. Stylohelia From. Jurassic ; Europe. abundant, Septa well numerous ; Jurassic, Family 5. Astraeidae i Milne Edwards and Haime. Corallum composite, or more rarely simple. Theca formed by fusion of sepfal edges. Septa numerous, usually well developed, upper edges toothed, serrated, or lobular ; visceral Chamber partitioned off by more or less abundantly developed dissepiments, more rarely by tabulae. Multiplication by budding or fission. Corallites of massive colonies usually reaching considerable altitude, and united with one another either directly by the walls or by means of septa exothecally produced (costal septa). Very abundant from thß Trias onwards, and by far the most protean family of all the Hexacoralla. According to the serrated or entire character of the free septal edges, Milne Edwards and Haime divided their Astraeidae into two subf amilies — the Astraeinae and the Eusmiliinae, the latter of which has been elevated by Yerrill to family rank. a. Simple coralla. Montlivaltia Lamx. (Fig. 138). Cylindrical, conical, turbinate, or dis- coidal, and either acutely pointed, or broadly expanded at the base. Septa numerous, upper edges serrated. Columella absent ; epitheca thick, corrugated, ^ The family name Astraeidae is not available for use among corals, as the generic name Astraea was applied by Bolten in 1798 to mollusks now referred to T^i7'bo and Xenophora, three years previous to its application, in 1801, by Lamarck to corals. It is known that the Astraeidae of Milne Edwards and Haime does not represent a natural association of corals, and, therefore, must be dismembered and divided into a number of families. Several subdivisions have already been proposed, but the detailed investigation of all the constituent genera has not progressed far enough to determine their natural affinities in all cases. In view of this condition it seems better to continue temporarily the use of the term Astraeidae until all the corals included under it have been thoroughly studied and their systematic affinities ascertained than to propose a Substitute name for one known to be invalid. .SUBCLASS II HEXACORALLA 1)7 readily becoming detached. Common in Triassic and Jurassic ; somewhat rare in Cretaceous and Tertiary. The genus should probably be made to include various species which have been referred to Epismilia From., and the so-called Oppelismilia Duncan. ß. Simple cor a IIa or composite coJonies iiiiiUiphjing hy calicinal or marginal gem- ination. Sfijlophyllum Eeuss. Corallum simple, either with or without calicinal or marginal gemmation, or f orming massive colonies. Septa stout, but only in- feriorly complete, terminating above in strong vertical spines. Dissepiments vesicular ; wall covered with epitheca. Alpine Trias. Stylophyllopsis Frech. Simple or imperf.ectly branching. Septa terminating near the centre in detached vertical spines. Alpine Trias. Fifi. 138. MontltvdUia caryophyllata (Lamx.). Great Oolite : Caen, Calvados. Natural size. Stylocora exilis Reuss. Mio- ccne ; Niederleis, Austria. a, Corallum in natural size. h, Calice enlarged (after Reuss). Bushy colonies multiplying hy lateral gemmation. Cladocora Ehrbg. Corallum composed of long cylindrical branches, free on all sides. Calice circular ; septa well developed ; columella papillous ; cycle of pali present. Jurassic to Recent. Stylocora Reuss (Fig. 139). Branches cylindrical; septa stout, those of the first cycle with columnar thickenings or inner edges ; columella styliform. Cretaceous and Miocene. Pleürocora E. and H. Cretaceous. Goniocora E. and H. Triassic and Jurassic. h. Composite corallites multiplying hy hasal gemmation ; huds arising from stolons or hasal expansions. Rhizangia E. and H. (Fig. 140). Corallites united by short, sub-cylindrical Fig. 140. Rhizangia michclini Reuss. Middle Cretaceous ; Gosau Valley, Austria. Natural size (after Reuss). Fio, 141. Viadang iaconferki Reuss. Miocene; Bischofs- wart, Moravia. a, Corallum, natural size ; b, Calice enlarged (after Reuss). Stolons. Calices shallow, circular; columella papillous. Cretaceous and Tertiary. Latusastrea d'Orbigny. Corallites arising from common basal expansion, Short and strongly inclined to one side, so that the calices acquire a semi- VOL. I H 98 COELENTERATA— ANTHOZOA PHYLUM II circular contour and assume the form of protruded lips. Jurassic and Cretaceous. Astrangia, Cryptangia, PhyUangia, Cladangia (Fig. 141), Ulangia E. and H., etc. Tertiary and Recent. €. Massive coralla muUiplying hy lateral gemmation. Orbicella Dana (Fig. 142). Cylindrical corallites united by exothecally produced, confluent, costal septa. Columella spongy ; dissepiments numerous between the septa both. within and exterior to the theca. Jurassic to Recent. Plesiastraea From. Like the preceding, but with several pali in front of all the cycles excepting the last. Tertiary and Recent. Fig. 143. Isastrea helianthoides (Goldf.). Coral-Rag ; Nattheim, Swabia. Natural size. Fig. 145. Favia caryophylloidcs From. Coral-Rag; Natt- heim. Natural size. Fig. 142. OrMcella conoidea (Reuss.). Mio- cene; Enzesfeld, near Vienna. a, Corallum, natural size ; b, Calices enlarged. Fig. 144. Latomeandra seriata Beck. Coral-Rag; Natt- heim. Natural size (after Becker), Fig. 146. CalamopJiyllia stokesi E. and H. Coral-Rag ; Steeple Ashton, England. Natural size. Isastrea E. and H. (Fig. 143). Corallites prismatic, closely crowded, and with fused walls. Calices polygonal ; columella imperf ect or absent. Trias to Cretaceous. Latortuandra d'Orb. (Fig. 144). Like the preceding, but with the calices situated in short furrows. Trias to Cretaceous. Stylastraea From. Lias ; Europe. Amphiastraea From. Upper Jurassic ; Europe. Leptastrea, Solenastrea, Prionastrea, E. and H. etc. Tertiary and Recent. {■. Massive coralla multiplying hy fission. Favia Oken (Fig. 145). Corallum massive; calices oval or distorted, and united by confluent costal septa ; columella spongy. Jurassic to Recent.' >UBCLASS II HEXACORALLA 99 Goniastrea E. and H. Corallites prismatic, calices polygonal. Septa well cleveloped ; columella spongy ; pali in front of all cycles excepting the last. Cretaceous to Recent. '7], Branching coralla multiplying hy fission. Calamophyllia Blainv. (Bhabdophyllia E. and H.; Lithodendron p. p. Mich.) (Fig. 146). Colony fasciculate or bushy ; corallites very long, cylindrical. Wall costate, without epitheca ; columella absent. Trias, Jurassic and Tertiary. Especially common in Alpine Trias. C. clathrata (Emmrich). Thecosmilia E. and H. (Fig. 147). Colony bushy, calices dividing by fission, and more or less free. Epitheca corrugated, readily wearing away ; columella absent or rudimentary. Trias to Tertiary. Accord- ing to Frech identical with Calamophyllia. Very common in Triassic and Jurassic. Baryphyllia From. HymenophyUia E. and H., etc. Cretaceous. Fig. 147. Thecosmilia tric (Goldf.). Coral-Rag heim. Natural size. hotoma : Natt- 0. Coralla with confluent calices increasing by fission. Leptoria E. and H. (Fig. 148). Corallum massive, composed of labyrinthic rows of confluent corallites with fused walls. Septa closely crowded, approaching parallelism ; columella lamellar. Jurassic to Tertiary. Diploria E. and H. Like the preceding, but with corallites united by produced costal septa instead of directly by their walls. Cretaceous to Recent. Aspidiscus König {F\g. 149). Corallum discoidal, circular or elliptical, covered on lower side with wrinkled epitheca. Cali- cinal furrows radiating from the centre outwards, and separated from one another by sharply crested ridges. In the centri- fugally disposed corallites the outermost septa are thickened, Cretaceous, Cretaceous. Symphyllia E. and Fig. 148. Leptoria koniiicki Reuss. Upper Cretaceous ; Gosau Valley. Natural Fig. 149. Aspidiscus cristatus König. Middle Cre- taceous ; Batiia, Al- geria. Natural size. and form by their union a banded margin. Stiboria Etall. Jurassic. Stelloria d'Orb. H. Tertiary and Recent. Family 6. Eusmiliidae Verrill. Like the Astraeidae, except that upper septal edges are entire, not serrated. a. Simple coralla. Trochosmilia E. and H. (Fig. 150). Turbinate, base acutely pointed or Septa numerous, extending to the centre. Without epitheca, costae granulated. Columella absent, dissepiments numerous. Cretaceous and Tertiary. encrusting, 100 COELENTERATA— ANTHOZOA THYLUM II Coelosmilia E. and H. (Fig. 151). Like the preceding, but with dissepi- ments sparsely developed. Cretaceous and Kecent. Placosmilia E. and H. (Fig. 152). Cuneiform, base acutely pointed or slightly pedunculate. Calice laterally compressed, elongated. Septa numerous ; Fig. 150. Trochosmilia granifera Haime. Turoiiian ; Bains-de- Rennes, France, a, Profile ; b, Calice slightly enlarged (after Fromentel). Fig. 152. Fig. 151. Coelosmilia laxa E. and H. White Placosmilia cuneiformis E. Chalk ; Lüneburg, and H. Upper Cretaceous ; Hanover. Natural St. Gilgen on Wolfgangsee, size. Anstria. Natural size. dissepiments abundant ; columella foliaceous. Epitheca absent ; costae granu- lated. Cretaceous. Diplodenium Goldf. Calice laterally compressed, greatly elongated in transverse direction, and bent downwards at the ends so as to become crescent- shaped. Columella and epitheca absent. Costae dichotomously or tricho- tomously furcate. Upper Cretaceous. Axosmüia E. and H. Jurassic. Fhyllosmilia From. Cretaceous. LopJio- smilia E. and H. Cretaceous and Eecent. ß. Coralla muUiplying hy lateral gemmation. Placophyllia d'Orb. (Fig. 153). Buds originating on calicinal margin or sides, and giving rise to bushy or massive colonies. Columella styliform Jurassic. Fig. 153. Placophyllia dlanthus (Goldf.). Coral- Rag ; Nattheim. a, Corallum, natural size ; b, Calice enlarged. Stylina delabechei E. and H. Coral-Rag ; Steeple Ashton, England, a, Natural size ; b, Calices en- larged. Galaxea Oken. Bushy colonies with cylindrical corallites united by layers of finely vesicular coenenchyma. Recent. Stylina Lam. (Fig. 154). Massive colonies, with corallites united by coalescent costae. Septa well developed, disposed in six, eight, or ten cycles. Dissepiments numerous ; columella styliform. Multiplication by costal gemma- tion. Profuse in Trias, Jurassic, and Cretaceous. SUBCLASS II HEXACORALLA 101 Cryptocoenia E. and H. Jurassic and Cretaceous. Massive colonies, with corallites united by costae. centre : colnmella absent. Visceral Chamber Astrocoenia deeaphylla B. and H. Upper Cretaceous ; Gosau Valley, Austria. «, Corallum, natural size ; b, Caiices enlarged. Placocoenia d'Orb. Cyathopliora Mich Septa short, not reaching the partitioned off by horizontal tabulae. Jurassic . and Cre- taceous. Coccophyllnm Reuss. Massive colonies, with corallites united directly by their walls. Caiices polygonal, septa numerous. Col- umella absent ; visceral Cham- ber tabulated. Alpine Trias. Pinacophyllum Frech. Tri- assic. Holocystis Lonsd. Massive colonies, with corallites united by costae. Four of the septa larger or stouter than the rest. Tabulae in visceral Chamber. Cretaceous. Astrocoenia E. and H. (Fig. 155). Massive colonies. Corallites polygonal, united by their walls ; septa numerous, long. Columella styliform ; only dissepiments present in visceral Chamber. Trias to Tertiary. Stephanocoenia E. and H. Like the preceding, but with columella surrounded by cycle of pali. Trias to Recent. Phyllocoenia E. and H. {Confusastrea d'Orb. ; Adelastrea Reuss). Massive colonies. Corallites round or oval, imperfectly united by costae. Septa strongly developed, thickened in the middle between theca and the centre. Columella rudimentary. Trias to Tertiary. Convexastrea d'Orb. Trias to Cre- ^Tliaa^ taceous. Columnastrea, Stylocoenia E. and H., etc. Cretaceous and Tertiary. y. Coralla multiplying by fission. Haplosmilia d'Orb. Bushy colonies. Corallites usually with dichotomously dividing crests. Caiices circular or elongated ; columella styliform ; theca with ridge-like costae. Jurassic. Plocophyllia Reuss (Fig. 156). Branching, foliaceous, or massive colonies. Corallites either becoming free or grouped into detached rows. Columella absent. Tertiary. Barysmilia E. and H. Corallum massive, forming a thick stem, the apex of which is covered with short buds. Caiices oval, sometimes disposed in series ; columella rudimentary. Cretaceous. Stenosmilia From. Like the preceding, but with lamellar columella. Cretaceous. Pachygyra E. and H. Corallites arranged in winding rows, and united by a broad mass of costal coenenchyma. Columella lamellar. Jurassic and Cretaceous. Fig. 156. Plocophyllia calyculata Reuss. Oligocene ; Monte Carlotta, near Vicenza. Natural size. 102 COELENTERATA— ANTHOZOA PHYLUM II Phytogyra d'Orb. Jurassic and Cretaceous ; Europe. Bhipidogyra E. and H. (Fig. 157). Corallum fan-shaped, often corrugated, and with but a single calicular furrow. Columella lamellar. Jurassic and Cretaceous. Fig. 157. Bhipidogyra crassa From. Coral-Rag ; Gray, Haiite-Saone. 1/2 natural size. Suborder 2. FUNGIDA Duncan. Solitary or colonial corals. Synapticulae in tlie interseptal and mtercostal loculi. Dissepiments present or absent. Septa lamellate and solid or slightly perforate, or composed of a trabecidar lattice-work with numerous perforations. Basal structures perforate or imperforate. Family 1. Fungiidae Dana. Embryo after becoming attached forms a trophozooid, which gives rise to buds (anthoblasts) ; these become detached, forming free individuals (anthocyathi). Adult corallum, simple or colonial, depressed or mitroid in form. Septa of higher cycles perforate, those of the lower perforate or solid. Synapticula, but no dissepiments present. Wall usually perforate in young, free individuals, subsequently more or less compact. No epitheca. Fungia Lamarck ; Halomiira Dana ; Polyphyllia Quoy and Gaimard ; Zoopilus Dana ; Cryptabacia E. and H. ; Lithactinia Lesson ; Herpetolitha Escholtz. Recent. Fungia oc- curs also in the post-Pliocene. Family 2. Agariciidae Verrill. Simple or colonial Fungids with lamellar, usually imper- forate, septa. Wall solid in simple genera, basal wall solid in colonies, walls between coral- lites solid when developed. Dis- sepiments present or absent. Microseris From. (Fig. 158). Corallum simple, dis- coidal, circular ; upper side vaulted, lower flat and granu- lated. Cretaceous. Trochoseris E. and H. Simple species, trochoid and fixed. Tertiary and Recent. Cyathoseris K Sind H. (Fig. 159). Corallum turbinate, attached. Young Fig. 158. Microseris hemis-phaerica From. Greensand (Cenomanian) ; Le Mans, France, a and b, Upper and lower surfaces, enlarged ; c, Profile, natural size. I.,.„ HEXACOKALLA 103 corallites arising from periphery by costal gemmation. Common outer wall naked, striated. Cretaceous and Tertiary. Leptophyllia Reuss (Fig. 1 60). Corallum simple, conical or cylindro-conicäl, with superficial calice. Septa numerous, thin, regularly-toothed ; solid or only Cyathoseris siCbregxdaris Keuss. Oligocene ; Monte Carlotta, near Viceiiza, Italy. a, Top view ; h, Side view, natural size. Fig. 160. Leptophyllia sinuosa From. Neocomian ; St. Dizier, Haute - Marne. Natural size. partially perforate. Thin dissepiments present. Thin epitheca present. Thamnasteria Lesauv. {Thamnastraea auct., of which T. lamourouxi is the type.) Jura. Lophoseris E. and H. ; Agarkia Lam. ; Siderastrea Blv., etc. Tertiary and Eecent. Family 3. Anabaciidae Duncan. (Pseudoasfraeinae and Pseudoagaricinae Pratz ; Microsolenidae Gregory.) Simple coralla, or composite, hasally expanded or massive colonies. Septa numerous, perforate, and composed of calcareous hodies (traheculae) arranged in vertical or fan-shaped rows (trabeculate). Theca hetween individual corallites ahsent, hut may he present on under side of corallites or on lower side of the common stock. Interseptal loculi with synapticula and dissepiments. Abundant from Trias to Cretaceous ; rarer in Tertiary and Recent. Änahacia d'Orb. Simple, free, discoidal, or lenticular coralla, with flat base. Upper side vaulted, calice slit-like. Septa very numerous, thin, and united by synapticulae. Theca absent. Jurassic ; Europe. Genabacia E. and H. Like the preceding but composite, the central calice being surrounded by a row of smaller calices. Jurassic ; Europe. Micrabacia E. and H. Cretaceous ; Europe. Omphalophyllia Laube. Simple, turbinate or sub-cylindrical, attached, and covered with epitheca. Septa very numerous, upper edges granulated. Calice shallow, columella styliform. Alpine Trias. Cyclolites Lam. (Fig. 161). Simple, free, discoidal, upper side vaulted, lower flat and covered with corrugated epitheca. Septa very thin, extending to the centre, extremely numerous, composed of vertical rows of traheculae, and united by synapticulae and dissepiments. Very abundant in Cretaceous, rare in Jurassic and Eocene. Dimorpharaea From. (Fig. 162, A, B). Composite, laterally expanded and pedunculate, or mushroom-shaped coralla. Common wall restricted to lower 104 COELENTERATA— ANTHOZOA PHYLUM II side of corallum ; individual corallites without proper walls, but united by costal septa. Columella styliform or rudimentary. Septa well developed, composed of fan-shaped rows of cylindrical trabeculae, and united by synapti- culae and dissepiments. Very abundant from Trias to Oligocene. Cyclolites undulata Lam. Fig. IGl. Upper Cretaceous ; Gosau Valley, Salzkainmergut. surface ; c, Lateral aspect of septum, natural size. (I, Side view ; h, Lower Dimorphastrea d'Orb. Like the preceding, but with calices concentrically arranged about a central individual. Trias to Tertiary. Co7noseris d'Orb. (Fig. 163). Like Dimorpharaea, but with calices separated into groups by ascending flexuous ridges. Jurassic and Tertiary. Astraeomorpha Eeuss. Coralla composite, tuberous, basally expanded, or branching, and covered with corrugated epitheca. Corallites small, united by Short and stout costal septa ; columella styliform. Trias to Oligocene. FKi. 162. A, Dimorplmraea, Lateral surface of costal septum, enlarged (8/1), showing ftabecular Constitution ; m, Line of junction of two septa belonging to difterent corallites ; t, Trabeculae ; p, Trabecular lacunae (after Pratz). B, Dimorpharaea agari- cites (Goldfuss). Upper Cretaceous ; Gosau, Salzkainmergut. Portion of upper surface of corallum, natural size. Fi(i. 163. Comoseris conferta Reuss. Oligocene ; Monte Carlotta, near Vicenza. Twice enlarged. Microsolena Lamx. Colony massive, polymorphous, mammiliform, conical, with a broad base, nearly spherical, turbinate and pedunculate. Trias and Jurassic. Suborder 3. PERFORATA Milne Edwards and Halme. Skeleton huilt up of small calcareous hodies (sclerites), between which are empty interstices of greater or lesser size. Theca formed hj fusion of outer septal edges or absent. Interseptal loculi empty throughout or traversed by synapticida or dissepiments. SÜBCLASS II HEXACORALLA 10^ Family 1. Archaeocyathidae Walcott.i Simple, turhinate, or sub-cylindrical coralla. Septa and theca porous ; inner septal edges united hy perforated interiar wall, which encloses a hollow central space. Synapticula present in interseptal loculi. All biit one of the genera described up to the present time (Archaeocyathus Bill., Ethmophyllum Meek, Spirocyathus Hinde, Protopharetra Bornem., etc.) occur in the Cambrian rocks of North America, Spain, Sardinia and Aus- tralia. Aükokania Walcott is known from the Lower Huronian of Ontario. They represent possibly a distinct order of the Madreporaria. Family 2. Eupsammidae Miliie Edwards and Haime. Corallum simple or hecoming composite hy lateral gemmation. Septa very numerous, sometimes united hy synapticula, and frequently witli their inner edges fused together. Theca naked or covered with epitheca, and formed hy thickening of the septal edges. Silurian to Recent. Calostylis Linds. Corallum simple, sub-cylindrical, or composite and multiplying by lateral gemmation. Septa very numerous, of spongy con- sistency, and either fused together or united by synapticulae. Columella thick, spongy ; wall covered with epitheca. Silurian ; Gotland. This genus probably belongs to the Tetracoralla. Haplaraea Milasch. Simple, cylindrical coralla, with broad Fio. 164. Eupsammia t rochtformis (Pallas). Cal- caire Grossier ; Chaussy, iiear Paris. Natural size. I '^^^^»sf^ist Fig. 165. Eahuwphyllia sinuata Reuss. Oligocene , Waldböckelheiin, Prussia. a, Natural size; b, Number of septa enlarged. Fig. 166. Steplumophyllia elegans (Bronn). Pliocene ; Stazzano, near Modena, Italy. a and h, Upper and lower surfaces, enlarged ; c, Profile, natural encrusting base. Septa numerous, extending to the centre, perforated by large apertures, and sometimes fused together or united by synapticulae. Dissepiments also present, but no columella. Jurassic and Cretaceous. Eupsammia 'E. and H. (Fig. 164). Conical or turbinate, acutely pointed, free. Septa very numerous, arranged in five cycles, those of the last cycle stouter than the rest. Columella present or absent. Eocene to Recent. Balanophyllia Wood (Fig. 165). Simple, sub-cylindrical, attached by the base. Columella spongy ; septa closely crowded, partly fused together. Eocene to Recent. Stephanophyllia Mich. (Fig. 166). Simple, discoidal; base horizontal, calice 1 Billings E., Palaeozoic Fossils of Canada, i., 1861-65.— WWc-o^^, C. D., Bull. U.S. Geol. Survey, No. 30, 1886.— ßor7iema?iw, /. G., Versteinerungen des Canibrischen Systems von Sardinien. \%%Q.— Hinde, G. J., Quart. Journ. Geol. Soc, 1889, vol. x\v.—Lambe, L. M., Revision of the Genera and Species of Canadian Palaeozoic Corals. Geol. Surv. Canada, Contrib. to Canad. Palaeont., 1899, vol. \\.— Taylor, W. T. G., The Archaeocyathinae. Mem. R. Soc. S. Aust., 1910, vol. ii. 106 COELENTERATA— ANTHOZOA PHYLUM II circular. Septa numerous ; the six priiicipal septa extending to the centre, the remainder with fused inner edges. Cretaceous and Tertiary. ^^ Dendrophyllia Blv. (Fig. 167). Corallum branching, increasing by lateral gemmation. Calices oval ; septa numerous and slender, those of the last cycle extending to the spongy columella, and fused with the converging ends of shorter septa of preceding cycle. Tertiary and Recent. Lohopsammia, Stereopsammia Edw. and H. Eocene. Astroides E. and H. (Fig. 97). Kecent. Family 3. Poritidae Dana. Composite coralla composed of porous sderenchyma. Corallites small ; septa as a rule only moder- Fio. 167. ately numerous, sometimes repre- DendrophyUia elegans Duncan. Oligocene ; Brockenhurst, Seuted by TOWS of trabeculae OT England, a, Corallum, natural size ; b, Transverse section of 7 77 m. i 1 caiice, eniarged, iameUae. I fieca aosent. Subfamlly A. Spongiomorphinae Frech. Corallum composed of thicJc trabeculae and strengthened by horizontal synapticulae. Calices very imperfectly differentiated from coenenchyma, and without distinct septa. Dissepiments usually sparsely developed. Of the genera belonging to this subfamily, Spongiomorpha, Heptastylis and Stromatomorpha Frech, are found in the Alpine Trias (Rhaetic and Zlambach beds). These are all tuberous, composite coralla of extremely irregulär form. In Spongiomorpha and Heptastylis, six septa are indicated by somewhat regularly disposed columns of trabeculae ; and in the latter form these are bound together by synapticulae which are projected at equal altitudes, and form perforated horizontal storeys. In a b Stromatomorpha no radial arrange- ment of the trabecular septa exists. Palaeacis E. and H. (Sphenopo- terium Meek and Worth.), occurring in the Lower Carboniferous limestone of North America and Scotland, per- haps also belongs here. Fig. 168. Adinacis elegans Reuss. Upper Cretaceous ; Gosau Valley, Salzkaramergut. a, Upper surface, natural size ; h, Transverse section, eniarged ; c, Longitudinal section, eniarged (after Reuss). Subfamily B. Poritinae Milne Ed- wards and Haime. Septa not very numerous, well developed. Corallites united by their porous walls. Litharaea E. and H. (Fig. 169). Massive coralla. Calices sub-polygonal, septa generally in three cycles ; columella spongy. Eocene and Miocene. Rhodaraea E. and H. Massive coralla. Spurious walls of corallites thick ; pali prominent. Miocene and Recent. SÜBCLASS II HEXACORALLA 107 Porites Link (Fig. 170). Massive or branching coralla. Calices shallow, polygonal ; septa irregularly reticulated, usually twelve in number ; columella Litharaea ivehsterl (Bowerb.). Eocene ; Bracklesliam Bay, England, a, Coralluin, natiiral size ; h, Four calices enlarged. Fig. 170. Porites incrustans Reuss. Miocene ; Mo- ravia. a, Tranverse section ; b, Longi- tiidinal section. Both tigures highly magni- fied. papillous, surrounded by a single cycle of pali, the latter five or six in number, and not very distinct from the septal ends. Endotheca exists sparingly, and may be dissepimental or tabulate, or may be mere stereoplasm. Cretaceous to Recent. Thegenus b c Porites is one of the most im- a ^^^ . portant of existing reef-builders. -^^^f^s^^ ^ Subfamily C. Alveoporinae Verrill. Septa composed of detached traheculae, spines, or reticulated lamellae, Theca perforate. Vis- ceral Chamber with perforate tabulae. Alveopora Quoy and Gaim. (Fig. 171). Massive coralla. Calices small, polygonal. Septa represented by detached spinous processes. remotely situated. Tertiary and Recent. KonincUa E. and H. Cretaceous ; Europe. Fig. 171. a,Alveoporaspongiosa'Dm\&. Recent ; Fiji Islands. Longi- tudinal section of corallite showing perforate vvalls and tabulae; h, Alveopora rudis Reuss. Nummulitic liniestone ; Oberburg, Styria, Vi ; c, Calices, greatly enlarged. (Fig. a, after Dana ; h, after Reuss.) Tabulae sparsely developed, Family 4. Acroporidae Verrill. Cor)iposite, branching, lobate, foliaceous, or massive coralla with corallites embedded in a canaliculated and reticulated coenenchyma. Septa (6-24) compact, sometimes imperfectly developed. Two long septa often projected from opposite sides and meeting in the centre. The genus Äcropova Oken (Madrepora auct., non Madrepora Linn., 1758) (Fig. 172), is an important agent in the construction of existing coral reefs, and builds colonies sometimes of considerable size. It occurs sparsely in the fossil State in Tertiary strata of various regions. Actmacis d'Orb. (Fig. 168). Massive or branching coralla. Coenenchyma abundant, granulated ; septa stout, of nearly uniform proportions, columella papillous ; pali in front of all the septa. Cretaceous and Tertiary. 108 COELENTERATA— ANTHOZOA phylum ii Astreopora Blv. Massive coralla. Coenenchyma porous and on upper surface echinulate. Septa of dissimilar proportions ; columella and pali absent. Tertiary and Recent. Dendracis E. and H, ; Cryptaxis Reuss. Tertiary. Turbinaria Oken (Gemmipora Blv,). Corallum foliaceous. Coenenchyma tolerably compact and finely echinulate. Septa of similar proportions : columella spongy. Cretaceous to Recent. Range and Distribution of the Hexacoralla. The group Aporosa of the Hexacoralla begins as the Tetracoralla disappear, and develop a great variety of forms in the Trias, from the Mesozoic onward to the present day they have continued to play a leading part in the con- struction of coral-reefs. Of the f amilies constituting the Aporosa, the " Astraeidae " is by all odds the most important and most protean, in com- parison to which the Fungida, Stylo- phoridae, Pocilloporidae, Oculinidae and Turhinolidae fall into greatly subordinate rank. The other f amilies are all younger than the "^s/raßw/ae/'not beginninguntil the Jurassic, the Pocilloporidae, indeed, not until the Tertiary. Fig. 172. The Eupsammidae and Poritidae of Jcroporo ajifirZica (Duncan). Oligocene; Brocken- the Perforata OCCUr Sporadicallv in the hurst, England, a, Calices enlarged ; b, Longi- ^-i • t r~i ^ -r ^ •^ -. • tudinal section, greatly enlarged. bllurian and üarboniterOUS, whllc it IS UOt until the Trias that the Anabraciidae and Poritidae develop a large variety of forms ; from the Trias to the Tertiary, however, these genera continue to be important reef-builders. The Eupsam- midae attain their greatest development in the Tertiary and Recent, while the Acroporidae belong almost exclusively to the present period. Occasional isolated deep-sea forms are met with in most of the several geological periods, but the usual mode of occurrence of the Hexacoralla is associated in masses in coral limestones ; the limestones may be of very vari- able thicknesses, but as a rule are interstratified between deposits of distinctly littoral character. Ancient coral-reefs most nearly resemble modern fringing or barrier reefs, but not atolls, the origin of which is clearly dependent upon peculiar conditions. The St. Cassian, Zlambach and Rhaetic beds of the Alpine Trias contain large numbers of reef-building Hexacoralla; but the pure limestones and dolomites of the Alps, as well as the Trias outside the Alpine region, are frequently either almost or entirely destitute of coral remains. In the Lias, coral-reefs have been found in England, Luxemburg and Lorraine. Certain beds of the Dogger, usually of but meagre thickness, are occasionally charged with corals, as in Swabia, the Rhine Valley in Baden, the Swiss Jura, Normandy and England. Coral limestones are abundantly developed in the Upper Jurassic of the Jura Mountains in France and Switzer- land, in Lorraine, Southern Baden, Swabia (Nattheim, Blaubeuern), Bavaria (Kelheim), many places in France and England, as well as in the whole province of the Alps, Carpathians, Cevennes and Apennines ; here the uppermost I CLAss III ALCYONARIA 109 horizon (Tithonian) is especially characterised by their development. Reef corals are also greatly developed in the Cutch (Jurassic) series of India. In the Lower Cretaceous (Neocomian) coral-reefs are found in France (Haute -Marne and Yonne), Crimea and Mexico; while the Urgonian of Switzerland and the Bavarian Alps is occasionally charged with corals. In the Turonian and Senonian of the Alps (Gosau Beds), Pyrenees and the Provence, numerous coral-reefs occur, usually accompanied by Budistae ; elsewhere, how- ever, except in Holland (Maestricht) and Denmark (Faxoe), the Upper Cretaceous contains but a limited number of reef-building Hexacoralla. In the older Tertiary (Eocene and Oligocene) occurrences of coral-reefs are known on the northern and southern flanks of the Alps and Pyrenees, in Arabia, India, the West Indies, and in Georgia, Florida, Alabama, Mexico and Central America ; outside these areas their distribution is mostly sporadic. In the Miocene and Pliocene the true coral-reefs retreat more and more towards the equator (Red Sea, Java, Japan, Gulf of Mexico), while the Hexacoralla which persist in geologic formations within the temperate zone (Vienna Basin, Italy, Touraine) constitute but an insignificant feature of the general fauna. [Tlie foregoing sections on the Tetracorallä and Hexacoralla have been revised by Dr. T. Wayland Vaughan, of the United States National Museum at Washington. It sliould be observed tliat, in the present unsatisfactory state of our knowledge of these organisms, the Classification adopted in this work, although perhaps as good as any available, is tentative in character. — Editok.] Subclass 3. ALCYONARIA Milne Edwards. {Octacfinia Ehrenberg ; Oäocoralla Haeckel). Composite colonies, rarely simple polyps, the individuals provided with eight mesenterial folds and eight broad, pinnately fringed, or plumose tentacles, which form a Single cyde about the mouth, Hard skeletal elements are very generally developed in the Alcyonaria, being absent in comparatively few forms, and are remarkable for their manifold variety ; they occur either detached in the ectoderm and mesoderm, or are closely packed together at the base to form a horny or calcareous axis (sclero- basis), about which the polyps are distributed. Sometimes the calcareous bodies (sderodermites) form compact tubes which are periodically partitioned off into storeys with the upward growth of the animal. Reproduction is accomplished either sexually, or asexually by basal or lateral gemmation, rarely by fission. Only the calcareous parts are known in the fossil state, such as the solid axes, detached skeletal elements, tubes and composite coralla ; the horny structures are totally destroyed during fossilisation. The Alcyonaria make their appearance in the Ordovician, but rarely occur in great abundance. Family 1. Alcyonidae Milne Edwards and Haime. Fixed, fleshy, lobate, or ramose polyp Stocks (very rarely simple individuals), with echinulate or spicular calcareous bodies (sderodermites) occurring detached in the soft parts. Isolated sderodermites readily escape Observation, owing to their minute 110 COELENTERATA— ANTHOZOA PHYLUM II size and fragile Constitution. They have been detected as yet only by Pocta in the Upper Cretaceous strata near Laun, Bohemia. I Fig. 173. Graphularia de- sertorum Zitt. Nummulitic lime- stone (Bocene) ; Farafreh, Libyan Desert, Africa. a, Axis, natural size ; b, Section of same ; c, Striated surface, enlarged. Family 2. Pennatulidae Milne Edwards and Haime. Folyp Stocks with hase embedded in sand or mud, and with horny or calcareous sclerohase ; polyps dimorphic. Slender, round or quadrate calcareous axes referable to tbe Pennatulidae have been detected with certainty only in the Trias (Prographularia Frech.), Cretaceous (Pavonaria Cuv. ; Pennatulites and Palaeosceptron Cocchi ; Glyptosceptron Böhm), and Tertiary (Graphularia E. and H.) (Fig. 173). Family 3. Gorgonidae Milne Edwards and Haime. Fixed, branching or fan-skaped colonies, vnth horny or calcareous solid sclerohase, or with jointed axis composed of alternating horny and calcareous segments. All the genera possessing horny, flexible axes (Gorgonia, Rhipidogorgia, etc.) are perishable. Detached remains referable to Primnoa, Gorgonella and Firgidaria, the axes of whicli are composed of both horny and calcareous layers, have been described from the Tertiary. In the genus Isis the axis con- sists of cylindrical calcareous segments alternating with horny connecting joints. It is found fossil in the Tertiary, and has been reported also from the Cretaceous. The genus MoltJcia, occurring in the Upper Cretaceous, has cylindrical joints which are pitted with slight depressions indicating the position of branches. In the red or gem coral (Corallium Linn.) the axis is built up of spiniform sclerites, which are united by a fibro- crystalline calcareous matrix impregnated with organic matter. It occurs only rarely in the fossil state, but is known from the Cretaceous and Tertiary. Family 4. Tubiporidae Milne Edwards and Haime. Coralla composed of red-coloured parallel calcareous tuhes connected by horizontal plates. The cylindrical tubes of the recent Organ-pipe Coral (Tubi- pora) are composed of spiniform sclerites, which are united with one another directly in such manner as to enclose small hollow Spaces appearing superficially as pores. The connecting horizontal plates or floors are tra- versed by canals which communicate with the visceral Chambers of the tubes by means of numerous round openings. New corallites are budded from their upper surfaces. Unknown in fossil state. ^ Poita, Philipp, Sitzimgsber. Akad. Wiss. Wien, 1885, vol. xcii. SÜBCLASS III ALCYONARTA 111 Family 5. Helioporidae Moseley.^ Calcareous coraMa, composed of two series of tuhiform corallites ; the larger tubes (autopores) are embedded in a strongly developed coenenchyma made up of smaller tubes (siphonopores). Both autopores and siphonopores are dosely tabulate ; the autopores are provided with ridge-like pseudosepta, which, however, do not correspond numerkally with the tentacles. The affinities of the Helioporidae with the Alcyonaria were first pointed out by Moseley. The larger polyps inhabit the autopores, and are furnished with eight mesenterial folds and a crown of eight tentacles ; while the smaller polyps, which are without either tentacles or sexual organs, are lodged in the siphonopores. The skeleton is composed of calcareous trabeculae, the same as in the Hexacoralla, from whose centres of calcification radial fibres extend outwards in caespitose fashion. The siphonopores multiply by intermural gemmation, while the autopores are formed by the coalescence and fusion of a number of the siphonopores. Heliopora Blainv. (Fig. 174, A, B). Corallum massive or ramose; autopores with 12-25 slightly developed pseudosepta, and embedded in a coenen- chyma made up of smaller siphonopores : the latter are more closely tabulate than the autopores. Cretaceous to Recent. Polytremacis d'Orb. (Fig. 174, C). Like Heliopora, but pseudosepta much more strongly developed, sometimes reaching nearly to the centre. Cretaceous. Octotremacis Gregory {Polysolenia Reuss non Ehrenb.). Miocene ; Java. Fig. 1T4. Heliopora partschi (Reuss). Upper Cretaceous ; St. Gilgen on Wolfgangsee, Salzkammergut. J, Corallum, natural size. iJ, Por- tion of surface, enlarged. C, Polytremmis hlainvilleana Reuss. Upper Cretaceous ; Gosau, Salzkammergut. Vertical section, enlarged. Family 6. Heliolitidae Lindström.2 Corallum. massive, more rarely ramose, varying from spheroidal to flabellate, composed of tubulär or vesicular coenenchyma enclosing corallites in the form of large cylindrical and numerous smaller angular cells ; both the macro- and microcorallites with tabulae. Usually twelve septa present in the large cylindrical cells, but these are often rudimentary. No mural pores ; basal epitheca present. Silurian and Devonian. The genera assigned to this family exhibit in their general appearance, finer structure and manner of multiplication, considerable resemblance to Heliopora, with which they were associated by Moseley, Nicholson, Bourne, 1 Moseley, IL N., The Structure and Relations of Heliopora caerulea. Pliilos. Trans. Royal Society, 1877, vol. clxvi. — Bourne, G. C, On the Structure and Affinities of Heliopra caerulea. Ibid., 1895, vol. clxxxvi. pt. 1. 2 Lindström, G., Reniarks on the Heliolitidae. K. Svensk. Vetensk. Akad. Handl., 1899, vol. xxxii. — Klär, J. , Die Koralleufauna der Etage 5 des norwegischen Silursystems. Palaeontogr. 1899, vol. xlvi. — Idem, Revision der mittelsilurischen Heliolitiden, etc. Videnskabs-Selskabets Skrift. I. Classe, No. 10, 1903. 112 COELENTERATA— ANTHOZOA PHYLÜM II Gregory and others. Here, as in Heliopora, the autopores are formed by coalescence of numerous siphonopores in the coenenchyma. On the other hand the corallites in the Heliolitidae have well-developed walls of compact homo- geneous matter, and as a rule also tvvelve strong septa are present, sometimes reaching nearly to the centre of the cylindrical Chambers. On account of these differences the family has been separated from Heliopora, and some authors have proposed to associate them with certain Hexacoralla, or with the problematical Tabulata. Heliolites Dana (StelliporeUa Wenzel; Nicholsonia Kiär) (Fig. 175). Corallum massive, nodular or ramose. Autopores with twelve more or less strongly developed pseudosepta, though occasionally represented by rows of spinules, and frequently with central columella. Siphonopores without septa, and multlplying by fission or intermural gemmation. Abundant from Ordovician to Devonian. Plasmopora E. and H. (Diploepora Quenst.). Like Heliolites, but having walls of the siphonopores incomplete, and tabulae of contiguous tubes fused together so as to form a vesicular tissue. Ordovician to Devonian. iyv. mMi"' Heliolites porosa Goldfuss. Devonian ; enlarged Fig. 175. Eifel. ^, Corallum, natural size. iJ, Portion of outer sui 1 . C, Longitudinal section, enlarged. Profaraea E. and H. (Stylaraea E. and H., 7ion Seebach). Low incrusting corallites, with relatively little coenenchyma. Tabulae present in the autopores. According to Kiär, this form and Coccoseris Eichwald are closely related, and perhaps identical. Ordovician and Silurian ; Scandinavia and North America. Cosmiolithes Lindstr. Corallites thin, lamelliform. Coenenchymal pores thick-walled, not all of the same size, with concave or obliquely directed tabulae. Autopores with well - developed septa. Silurian ; G. ornatus Lindstr. Plasmoporella, Palaeoporites and Trochiscolithus Kiär. Silurian ; Scandinavia. Acantholithus, Pycnolithus Lindstr. Eelated genera having the same distri- bution. Appendix to the Anthozoa. Sub Order. TABULATA Milne Edwards and Halme. ^ Invaridbly composite coralla composed of tubiform or prismatic corallites. Walls thick, independently calcified, compact or perforated hy connecting mural pores. Septa ^ Literature : Lindström, O., Affinities of the Anthozoa Tabulata. Ann. Mag. Nat. Hist., 1876, ser. 4, vol. xviii. — Nicholson, H. A., On the Structure and Affinities of the Tabulate Corals of the Palaeozoic Period. London, 1879. — Roemer, F., Lethaea Palaeozoica, i., 1883, p. 416. — Waagen, suBCLAss III TABULATA— FAVOSITIDAE 113 hut slighüy developed (usualli/ six or twelve), sometimes 7'epresented merely hy vertical ridges or rows of spines, and sometimes entirely absent. Visceral Chamber partitioned ojf into successive storys by tabulae. Synapiiculae and dissepiments wanting. To the Tabulata were originally assigned by Milne Edwards and Haime all corals having numerous tabulae and rudimentary septa. Later researches have shown, however, that some of these forms (Focilloporidae) belong to the Aporosa, others (Ilelioporidae) to the Alcyonaria, and still others (MiUepora) to the Hydrozoa. The majority of the typical Tabulata (Favositidae, Syringoporidae, Halysitidae) exhibit close relationships to the Hexacoralla ; but since they are for the most part now extinct and are largely confined to the Paleozoic rocks, the determination of their systematic position is a matter of much difficulty. The ontogeny of the corallites in the Tabulata shows that the development of mural pores is homologous with the process of gemmation. Reproduction sometimes takes place by fission, but generally by means of buds from the edges of the calices at various stages during the growth of the parent corallites. Buds are given ofF early in Aulopora, producing basal corallites only ; periodically in Romingeria, producing verticils of corallites ; periodically and on one side in Halysites, producing linear series of adjacent corallites ; and very frequently in Favosites, etc., producing compact coralla with numerous mural pores representing aborted buds. Family 1. Favositidae Milne Edwards and Haime. Massive or branching coralla. Corallites uniformly prismatic, toll and united by their walls, which are perforated by large-sized pores. Septa very short, usually represented by but faintly projecting ridges or rows of spines, but seldom completely absent. Tabulae numerous, situated at regulär intervals, complete and horizontal, more rarely oblique or irregularly vesicular {^' cystoid"). The Favositidae are distinguished from the Poritidae, with which Verrill associates them, by their thick solid walls, which are perforated by round, sometimes tubiform mural pores. The corallites are usually polygonal in contour, and their walls exhibit in transverse sections a dark, or sometimes light-coloured median line, with thickenings of stereoplasma on either side (Fig. 176, C). The family is exclusively Paleozoic, and plays an important part in the formation of Silurian, Devonian and Carboniferous coralline limestones. Favosites Lam. {Calamopora Goldf.) (Fig. 176). Corallum massive, more rarely branching. Corallites prismatic, polygonal, generally hexagonal. Mural pores distributed at considerable intervals. Septa very faintly developed, represented by longitudinal ridges or rows of spines, or occasion- ally obsolete. Tabulae numerous. Ordovician to Carboniferous ; very abundant in Silurian and Devonian. W., and Wentzel, J., The Salt Range Fossils. Palaeontol. Indica, 18S7.—Beecker, C. U., The Development of a Palaeozoic Poriferous Coral. — Symiiietrical Cell Development in the Favositidae. Trans. Conn. Acad., 1891, vol. viii. — Wentzel, J., Zur K(3untniss der Zoantharia tabulata. Denkschr. Akad. "Wien, 1895, vol. \xü.—Sardeso7i, F. W., Über die Beziehungen der fossilen Tabulaten zu den Alcyonarien. Neues Jahrb. Mineral., 1896, Supplem. vol. x. — Weissermel, W., Sind die Tabulaten Vorläufer der Alcyonarien? Zeitschr. deutsch, geol. Ges., 1898, vol. 1. — Vinassa de Regny, P. E., Trias-Tabulaten, etc. Res. d. wisseusch. Erforsch, des Balaton-Sees, vol. i. pt. 1, Badapesth, 1901. VOL. I I 114 COELENTERATA— ANTHOZOA PHYLÜM II Columnopora Nich. {Calapoecia Billings). Like the preceding, but with numerous, short, well-marked septa. Mural pores large, disposed in vertical rows between the septa. Ordovician. Fig. 17(3. Favosites polymorpha (Goldf.). Devonian ; Eifel. A, Coralluni, natural size. B, CoraUites enlarged, two of them broken open and showing tabulae. C'and D, Transverse and longitudinal sections showing spiniform septa and mural pores {p). (C and B after Nicholson.) Emmonsia E. and Ordovician (Trenton). H. Ordovician to Carboniferous. Nyctopora Nich. Syringolites Hinde. Silurian (Niagara). Pachypora Lind. (Fig. 177). Corallum branch- «-— «^»^ ing, composed of prismatic, polygonal corallites, ^ ySKmJIm *^® walls of which are so thickened towards their VU/^Mf mouths by layers of stereoplasma that the calices wMyRjfc appear to have circular contours. Septa very minute ; mural pores scanty, but often of large size. Abundant in Silurian and Devonian. Trachypora E. and H. Dendroid with cylindrical stems. Corallites polygonal ; walls so thickened by layers of stereoplasma that the calices become round and greatly contracted, and appear to be superficially widely separated. Mural pores few and irregularly distributed. Septa represented by rows of spines. Tabulae at remote intervals. Common in Devonian. Striatopora Hall (Fig. 178). Like the preceding, but with tubes con- tracted by stereoplasma at a greater depth, so as to give the calices a funnel-shaped appearance. Silurian and Devonian. Alveolites Lam. (Fig. 179). Corallum massive or branching, composed of small, contiguous, compressed, thin-walled corallites, with obliquely opening triangulär or semilunar calices. Septa very faint, represented merely by ridges or rows of spinules, sometimes but a Single row present. Mural pores of large size, ir- regularly distributed. Very common in Silurian and Devonian. öladopora Hall ; Coenites Eichw. Silurian and Devonian. Pleurodidyum Goldf. (Fig. 180). Corallum depressed, discoidal, circular Fig. 177. Fachypora nicholsoni Frech. Middle Devonian; Eifel. A, Transverse section. B, Longi- tudinal section, enlarged ; p, Mural pores (after Nicholson). Fig. 178. Striatopora flexnosa Hall. Silurian (Niagara) ; New York. SUBCLASS III TABULATA— FAVOSITIDAE 115 or elliptical in contour, lower surface covered with concentrically striated epitheca, and frequently a foreign vermiform body occupying the ceiitre of the base. Corallites small, polygonal, contracted inferiorly so as to become funnel-shaped. Septa represented by faint marginal ridges, or obsolete. Fig. 179. A, Alveolites suborhicukiris L&m. Middle Devonian ; Gerolstein, Eifel. Natural size. B and (', Alveolites lahechei E. and H. Silurian (Wenlock) ; Ironbridge, England. Tangential and vertical sections, lo/j (after Nicholson). Fig. 180. Pleurodicty u m prohlemaiicum, Goldf. Lower Devonian ; Co- blenz. Natural size. Vermiform foreign body in the centre. Walls pierced by irregularly distributed mural pores ; tabulae sparse. Devonian. P. prohlematicum Goldfuss, is rather abundant in the Lower Devonian " Spirifera sandstone " of the Eifel, but is known only in the form of casts. In these the walls of the corallites are represented by narrow fissures which are bridged across by transverse rods, while the visceral Chamber is filled up with sandstone, P. stylopora Eaton, from the Hamilton Group of North America, is a closely related species and also possesses the vermiform body. Ä B C Fig. 181. Michelinia fuvosa de Kon. Carboniferous Limestone ; Tournay, Belgium. A, Corallum fronl above. B, Lower surface with radiciform epithecal processes. C, Vertical section (after Gaudry). Michelinia de Kon. (Fig. 181). Discoidal or hemispherical coralla, often of considerable size, and covered on the under surface with concentrically striated epitheca, which frequently develops hollow radiciform processes. Corallites polygonal, rather large. Septa represented by numerous longi- 116 COELENTERATA— ANTHOZOA FHYLUM II tudinal Striae or ridges ; mural pores irregularly distribiited ; tabulae very numerous, oblique or curved, incompletely developed, and usually filling the visceral Chamber with loose vesicular tissue. Devonian and Carboniferous. M. favosa de Kon., extraordinarily profuse in the Lower Carboniferous Lime- stone of Belgium. Family 2. Auloporidae Nicholson (Tuhulosa Milne Edwards and Haime). Creeping, hranching or reticulated tubulo,r coralla, composed of cylindrical, leaker or trumpet-shaped corallites, with thicJc, imperforate, wrinkled walls. Sepia repre- sented hy faint marginal Striae ; tahulae moderately numerous or wanting. Reprodudion by basal or lateral gemmation. Ordovician to Carboniferous. Aulopora Goldf. (Fig. 182). All the corallites of the prostrate corallum are attached by the whole of the lower surface to some foreign object (Alveolites, other corals, or mollusks). Tabulae more or less curved ; reproduction by basal gemmation. Ordovician to Carboniferous. Cladochonus M'Coy (Fyrgia E. and H.). Corallum branching, attached only at isolated points, and composed of funnel-shaped corallites without tabulae and septa. Reproduction by lateral gemmation. Carboniferous, Romingeria Nich. [Quenstedtia Rom.). Spreading, semi-erect, bushy coralla, only basally attached, and with cylindrical corallites increasing by lateral gemmation. Tabulae moderately numerous, horizontal. Silurian and Devonian. Fig. 182. A ulopora t ubaeformis Goldf. Devonian ; Gerol- stein, Eifel. Natural size (after Goldfuss). Family 3. Syringoporidae Milne Edwards and Haime. Fasciculate coralla composed of cylindrical corallites, unifed at intervals along the sides by hollow connecting processes or hy horizontal expansions. Walls thicJc, wrinkled; septa faintly developed, represented by delicate ridges or longitudinal rows of spinules ; tabulae numerously developed, usually irregularly funnel-shaped. Reproduction by basal gemmation or by buds arising from the connecting processes and horizontal expansions. Ordovician to Car- boniferous ; maximum in Devonian and Carboniferous. Syringopora Goldf. (Fig. 183). Fasciculate coralla, often attaining considerable size, and composed of cylindrical, thin-walled, somewhat flexuose corallites; the latter communicate by means of hollow, cylindrical, connecting processes. Septa rudimentary ; tabulae funnel-shaped. Corallum commencing with prostrate basal zooids similar to Aulopora. Numerous species ranging from Silurian to Carboniferous. Chonostegites E. and H. Corallum massive ; cylindrical corallites connected by horizontal, hollow, laminar expansions into which the endothecal tissues are directly continued ; tabulae oblique, cystoid. Devonian. Fig. 183. Syringopora ramulosa Goldf. Carboniferous Liinestone ; Regnitzlosau, Fichtelgebirge. Natural size. I Fia. 184., BCLAss III TABULATA— CHAETETIDAE 117 Thecostegites E. and H. Corallum encrusting ; corallites short, cylindrical, and connected by thick horizontal plates. Tabulae approximately horizontal ; septa twelve in number, represented by marginal ridges. Devonian. Family 4. Halysitidae Milne Edwards and Haime. Chain corals. Corallum composed of long, cylindrical, laterally compressed corallites, which are joined to one another only along the more restricted edges, and form free, vertical, intersecting and anas- tomosing laminae. Wall thick, covered on free sides by wrinUed epitheca ; tabulae numerous, horizontal or concave ; septa repre- sented by vertical ridges or rows of spines, in cycles of twelve, sometimes entirely absent. Increase by stolonal gemmation. The unique genus Halysites Fischer (Catenipora Lam.) (Fig. 184), comprises two groups of species ; those in which the corallum is composed throughout of coral- lites of equal size (H. escharoides Lam. sp.), and others in which any two of the larger corallites are separated by the intervention of a single smaller, closely tabulate tube , . , ^-r , . ^ . \/^i-. T ri", • . Halysites catemilana {II. catenmana L\nn. sp.) Urdovician and JSilunan ; maxi- (Linn.). Siiurian ; Got- mum in the Siiurian. ^^"^- Natural size. Family 5. Chaetetidae Milne Edwards and Haime. Massive coralla, composed of fine, subequal, tubiform corallites, contiguous on all sides. Calices rather irregulär in shape, one diameier slightly greater than the other. Walls thoroughly amalgamated, common to adjacent corallites, imperforate, apparently composed of closely arranged, anJcylosed vertical columns, which terminale at the surface in hollow prominences. Septa absent, but one or two tooth-like projections often observable in sections. Tabulae horizontal, remote or abundant. The forms belonging to this family are extinct, and occur chiefly in the Ordovician, Siiurian, Devonian and Carboniferous Systems ; but a few are also found sporadically in the Trias, Jura and Cretaceous. They are largely concerned in the formation of Paleozoic coral reefs, especially during the Carboniferous. Milne Edwards and Haime regarded them as Anthozoans, Rominger and Lindström as Bryozoans, while Dybowski emphasised their affinities with the Favositidae. By Nicholson they were assigned to the Octo- coralla, for the reason that the corallites frequently exhibit a dimorphous character the same as in Heliolites and Heliopora, besides agreeing in their microscopic structure with Heliolites; while in addition they possess well- developed tabulae and imperforate walls. Many genera and species formerly included under this family are now assigned to the Bryozoa. Chaetetes Fischer (Figs. 184«, 1846). Corallites long, thin-walled, prismatic, polygonal, all of one kind, and multiplying by fission. Uncompleted fission of the tubes often indicated in section by tooth-like projection extending into the visceral Chamber. Walls structureless, without dark median line ; tabulae complete, remote. Very abundant in Lower Carboniferous ; found 118 COELENTERATA— HYDROZOA PHYLUM II also in Lias and Upper Jurassic. C. radians Fisch., is an important rock- builder in the Russian Lower Car- boniferous, especially near Moscow. Dania E. and H. Silurian. Tetra- dium Dana. Ordovician and early Sil- urian. Pseudochaetetes Hang. Upper Jurassic; Europe. P. pohjporus (Qaenst.). Fig. 184a. Chaetetes septosus Fleiii. Lower Car- boniferous ; England. A, Transverse section parallel to npper surface. B, Vertical section both enlarged ; p, Pro- jecting spines representing uncompleted fission (after Nicholson). Fio. 184^. Chaetetes radians Fischer. Lower Carbon- iferous ; Moscow, Russia. Portion of longi- tudinally fractnred coralluni, natural size. Geological Range of the Tabulata. With but few exceptions the Tabulata are restricted to Paleozoic for- mations, and from the Ordovician to the Carboniferous Systems inclusive occur in considerable profusion, being associated with the Tetracoralla and certain Hydrozoa (Stromatoporoids) in the building of large coral reefs. Of the various families constituting this group, the systematic position of which is uncertain, the Halysitidae are limited to the Ordovician and Silurian, and the remainder, including the genus Chaetetes Fischer, are represented continuously from the Ordovician to the Carboniferous. In the Middle Cambrian shales of British Columbia, Walcott has recently discovered a remarkably well- preserved actinian, named Mackenzia, which appears to belong to the family Edwardsiidae, and to be closely related to the genus Edwardsia. Class 2. HYDROZOA Huxley. Hydroids and Medusae.i Sessile or free-swimming polyps or polyp Stocks, without oesophageal tube, and with simple gastrovascular cavity not divided into radial pouches. The Hydrozoans are organisms which rarely secrete hard parts, and hence are ill-adapted for preservation in the fossil state. The ramifying polyp Stocks are usually inferior in size to those of the Anthozoa, and possess always a simpler structure ; dimorphism or polymorphism is, however, exhibited by the difFerent individuals, some of which perform solely vegetative, and others only reproductive or protective functions. Of great interest is the prevailing alterna- tion of generations, in which process fixed polyp Stocks give rise to a generation of free-swimming Medusae, the eggs of which develop in turn into polyps. 1 Huxley, T. II., The Oceanic Hydrozoa. London, 1859. — Agassiz, A., North American Aca- lephae. 111. Cat, Museum Comp. Zool. Cambridge ii., 1865. — Ilincks, T., Natiiral History of the British Hydroid Zoophytes. London, 1868. — Claus, C, Untersuchungen über die Organisation, etc., der Medusen. Leipzic, 1883. I.^„ HYDROMEDUSAE 119 The Ilydrozoa are all aquatic, and with few excepfcions are inhabitants of the sea. They are commonly divided into the two following sub-classes : — Hydromedusae and Acalephae. Subclass 1. HYDROMEDUSAE Vogt. Sessile or free-swimming, usually hranching colonies, with dimorphic, nutritive and reprodudive polyps ; the latter frequently become liberated in the form of small, free- swimming Medusae, with non-lobate umbrellas composed of a hyaline, gelatinous substance. Six Orders af Hydromedusae are recognised : Hydrariae, Hydrocorallinae, Tubulariae, Campanidariae, Trachymedusae and Siphonophorae. Of these only the Hydrocorallinae, Tubulariae and Campanulariae secrete calcareous or chitinous structures capable of preservation in the fossil State. Order 2. HYDROCORALLINAE Moseley.i Naked polyps secreting at the base a dense calcareous skeleton, traversed at intervals by two series of vertical tubes, into which the dimorphic zoöids can be retracted. The Hydrocorallinae comprise the two Recent groups Milleporidae and Stylas- teridae, which were univereally regarded as true corals until Louis Agassiz and Moseley proved their relation- ship to the Hydrozoa. Millepora Linn. (Fig. 185). Massive, foliately expanded, encrusting or branching poly- paria (coenosteum), often attain- ing considerable size. Upper surface punctured by round openings of the larger tubes (gastropores), between which are the mouths of numerous smaller tubes (dactylopores). The skeleton is composed of a network of anastomosing calcareous fibres, traversed by a System of tortuous canals. The gastropores lodge the larger, nutritive polyps, and the dactylopores the smaller, food-procuring zooids ; the latter have no mouths, but are provided with short, clavate tentacles on their sides, and their tubes communicate with the vermiform canals. Zooidal tubes tabulate, but nonseptate. The genus is an important reef-builder of the present day, but occurs only sparsely in the fossil State. Earliest known forms appear in the Eocene. 1 Literature : Allman, J. G., Moiiograph of the Gymnoblastic or Tubularian Hydroids. Ray Society, 1871-72. — Moseley, II. N.. Philosopliical Transactions Royal Society, vol. 167, 1878, — Steinmann, G., Über fossile Hydrozoen aus der Familie der Coryniden. Palaeontographica, vol. XXV., 1877. — Idem, Über triasische Hydrozoen vom östlichen Balkan. Sitzber. Akad. Wiss. Wien, math. -phys. Classe, vol. cii.,1893. — Canavari, M., Idrozoi Titoniani appartenanti alla Faraiglia delle Ellipsactinidi. Mem. Comitato Geol. vol. iv., 1893. — Vinassa de Regny, 0. A'., Studii sulle Tdractinie fossili. Mem, Accad. dei Lincei, 1899, ser. 6, vol. iii. Fio. 185. Millepora nodosa Esp. Receiit. A, Upper surface of coeno- steum, showing gastropores, k, and dactylopores c, -lo/j. B, Vertioal section, k, gastropores with tabulae, t ; c, Vermiform canals communicating with dactylopores, 80/j (after Steinmann). 120 COELENTERATA— HYDEOZOA PHYLUM II Stylaster Gray. Branching polyparia composed of a network of fibrous, rose-coloured coenenchyma, in which are situated calicular depressions that are provided with pseudosepta and columellae, and communicate with the zooid tubes and vermiform canals. Recent, and occurring sparsely in the Tertiary. Order 3. TUBULARIAB AUman. Polyp Stocks which are either nahed or covered with chitinous outer layer (periderm). Both the polypoid nutritive zooids, and also the medusoid reproductive animals are without cup-shaped hydrothecae surrounding the polyp head. A chitinous or calcareous skeleton (hydrophyton) is frequently secreted at the hase. Hydractinia v. Bened. (Fig. 186). Hydrophyton in the form of encrusting, chitinous, rarely calcareous expansion, frequently investing gastropod shells. The crust consists of successive, slightly separated, horizontal laminae, which l ^- are supported by num- erous vertical rods or columns {radial pillars). The surface is covered with projecting hollow spines and tubercles, and is also traversed by shallow, branching grooves {astrorhizae). Interlaminar spaces communicating with the surface by means of rounded tubes. Tertiary and Recent. Ellipsactinia Steinm. Hydrophyton irregu- larly ellipsoidal, com- posed of thick, con- centric, slightly separ- ated, calcareous lam- ellae, which are united by sparsely distributed vertical columns. Lamellae are formed by the anastomosis of exceedingly delicate calcareous fibres, punctured by numerous fine radial tubes, and furnished on both sides with pits, tubercles and branching furrows. Upper Jurassic (Tithonian) ; Alps, Carpathians and Apennines. Sphaeractinia Steinm. Like the preceding, but composed of thin, widely separated lamellae, which are supported by numerous radial pillars. Centre frequently occupied by a foreign body. Upper Jurassic (Tithonian). Loftusia Brady (Fig. 187). Ellipsoidal or fusiform bodies, composed of thin, concentric or spirally rolled calcareous lamellae. Interlaminar spaces wide, intersected by numerous radial pillars, and often secondarily filled with calcareous mud. Eocene ; Persia. Parkeria Carp. Globular or walnut - shaped organisms with nodulated Fig. 186. A, Hydractinia echinata F\em. Recent; North Sea. Portion of commensal colony, greatly enlarged ; hy, Polyps (hydranths) ; go, Generative buds (gono- phores) ; hph, Hydrophyton adherent to shell of Buccinium undatum, and showing reticulated structure in vertical section. B, Hydractinia calcarea Gart. Vertical section of hydrophyton, greatly enlarged (after Carter) ; a, Primary basal lainella ; h, Interlaminar space ; c, Second lamella ; d, Radial pillars between the lamellae ; e, f, Tubercles and spines projecting on upper snrface. C, Hydractinia pliocaena Allm. Pliocene ; Asti, Italy. Hydro- phyton encrusting on Nassa shell (natural size). D, Portion of magnified surface of the latter, showing branching grooves and wart-like tubercles. CLASS II TUBULARIAE 121 exterior, and composed of rather thick, concentric, calcareous lamellae. Inter- laminar Spaces divided into chamberlets by stout radial pillars, which usually extend continuously through a number of lamellae. Both lamellae and pillars consistof minutely tabulated tissue, the tubules of which are radial in arrange- ment. Centre frequently occupied by a foreign body. Cambridge Greensaiid (Cenomanian). The genera Parkeria and Loftusia were originally described as agglutinated Foraminifera ; they are, however, manifestly very closely allied to Ellipsaäinia and Sphaeractinia. Porosphaera Steinm. (Fig. 188). Globular masses of the size of peas or hazel-nuts frequently growing around some foreign body, and composed of anastomosing calcareous fibres which are penetrated by numerous radial tubules ; the latter open on the surface in the form of large pores, around which radial or stel- late furrows (astrorhizae) are sometimes grouped. Upper Cretaceous. Fio. 187. Fig. 188. Porosphaera glöbularis (Phill.). Upper Cretace- ous ; Rügen. A, Skeleton, natural size ; l, Cavity originally occupied by foreign body, B, Trans- verse section showing Loftusia persiraBrady. Eocene ; Persia. yl, Specimen cut open to showgeneral radial tubes of gastro- structure, natural size (after Brady). B, Section showingitwo lamellae and inter- pores, 2/j (after Stein- laminar Alling, greatly enlarged. mann). Stoliczkaria Duncan. Trias ; Karakoram and Balkan Mountains. Cydoporidium, PMzoporidium Parona. Cretaceous. Poraäinia and Cyclac- tinia (Kerunia Mayer-Eymar) Yinassa. Tertiary. Heterastridium Reuss (Syringosphaeria Duncan). Includes spheroidal, nodular bodies of considerable size, composed of slender, anastomosing, and more or less distinctly radial calcareous fibres. Skeleton comparatively dense, but perforated by two series of zooidal tubes appearing superficially as pores. The apertures of the larger tubes are round, those of the smaller stellate, and are surrounded by radial furrows. Alpine Trias. Appendix to the Hydrocorallinae and Tubulariae. Stromatoporoidea Nicholson and Murie.^ Closely allied to the Hydrocorallinae and Hydractinia are the extinct Stromato- poroidea, which combine in many respects the characters pertaining to both of ^ Bargatzki, A., Die Stromatoporen des rheinischen Devons. Bonn, 1881. — Nicholson, H. A., Monograph of the British Stromatoporoids. Palaeont. Soc. 1886-92.— (Jir^y, G. H., Revision of the Sponges andCoelenterates of the Lower Helderberg Group of New York. 14th Ann. Rept. N.Y. State Geol., I^'d^. — Tornquist, A., Über mesozoische Stroniatoporiden. Sitzber. preuss. Akad. Wiss., 1901, vol. xlvii. — Yabe, H., On a Mesozoic Stromatopora. Jouru. Geol, Soc. Tokyo, 1903, vol. X. — Beninger, K., Einige neue Tabulaten und Hydrozoen aus mesozoischen Ablagerungen. Neues Jahrb. f. Mineral, i., 190Q.—Rothpletz, A., Über Algen und Hydrozoen im Silur von 122 COELENTERATA— HYDEOZOA PHYLUM II the above-named groups, but whose exact position in the zoological System remains as yet uncertain. During the Paleozoic era, to which they are con- fined, the Stromatoporoids were important geological agents, whole beds of limestone being often essentially constituted of their remains. In the Mesozoic era they are replaced by very closely allied forms of Hydradinia, which in all probability represent their immediate descendants. The Stromatoporoids secrete hemispherical, globiüar, nodular or hori- zontally expanded skeletons, which are sometimes encrusting, sometimes attached by a short basal peduncle, and are covered on the under side with concentrically wrinkled epitheca, while the apertures for the emission of the polyps are situated on the upper surface. The general tissue of the coenosteum is composed of numerous, concentric, undulating, calcareous layers or laminae, which are separated by narrower or wider interlaminar Spaces, but are at the same time connected by numerous vertical rods {radial pillars). The pillars as well as the laminae are traversed, as a rule, by minute, irregularly directed j ß canaliculi. In some genera the coenosteum is provided with verti- cal tabulate tubes, which most probably served for the recep- tion of the polyps, as in the genus Millepora ; but in many instances they are wanting. The surfaces of the laminae typically exhibit pores and small tubercles, and frequently also shallowstellate furrows (astrorhizae), which radiate outwards from numerous centres. Sometimes the laminae consist merely of a loose network of horizontal calcareous fibres. Goldfuss at first held the Stromatoporoids occurring so profusely in the Eifel for corals (Millepora), and subsequently for sponge-like zoophytes ; while von Rosen considered them as horny sponges that had become secondarily calcified. Sandberger and F. Roemer assigned them to the Bryozoans ; Dawson to the Foraminifera ; Sollas to the siliceous sponges {Hexacünellida) ; and Salter to the calcareous sponges, whose example Nicholson also followed. Lindström, Carter and Steinmann subsequently pointed out their relations to Hydradinia and Millepora ; and Nicholson and Murie came finally to regard them as a group of extinct Hydrozoans allied to Hydradinia on the one band (Actinostromidae), and Millepora on the other (Stromatoporidae and Idiostromidae). Adinostroma Nich. (Fig. 189). Skeleton having vertical or radial pillars disposed at tolerably regulär intervals, and extending continuously through all Gotland und Ösel. K. Svenslc. Vetensk. Akad. Handl., 1908, vol. xliii. — Gürich, G., Les Spongio- stromides du Visedansle Province de Namur. Mein. Mus. Roy. d'Hist. Nat. Belg., 1906, vol. iii. — Idem, Neues Jahrb. f. Mineral., 1907, i. — Parona, C. F., La Fauna coralligena del Cretaceo dei Monti d' Ocre nell' Abruzzo Aquilano. Mem. Conim. Geol. Ital., 1909, vol. w.— Parks, W.^A., The Stromatoporoids of the Guelph Formation in Ontario ; the Niagara ; the Silurian of America ; the Ordovician. Univ. of Toronto Studies, Nos. 4-7, 1907-1910. Fig. 189. Actinostroma intertextum Nich. Silurian (Wenlock) ; Shropshire. A, Tangential section showing radial pillars and reticulated structure of con- centric laminae. B, Vertical section, showing formation of laminae from processes given off horizontally by radial pillars, 12/1 (after Nicholson). |^p>ASS II STROMATOPOROIDEA 123 or at least a considerable number of laminae ; in vertical sections, accordingly, exhibiting a quadrate meshwork. The laminae consist of an anastomosing network of calcareous fibres, generally having a porous structure; their surfaces are covered with projecting granules or tubercles, which represent the free Upper ends of the vertical pillars. Rare in the Silurian, but very abundant Fic. l'JO. Stromatopora tuberculata Nich. (Corniferous limestone) ; Jarvis, Natural size (after Nicholson). Devon ian Ontario, Fin. 191. Cmmopora placenta Phill. Devonian ; Torquay, Devonshire. A, Tangential section, natural size. B, The same, highly magnified ; a, Vertical " Cauno- pora tube " ; h, Canal partially cut into ; c, Calcare- ous fibres traversed by delicate ramifying canaliculi. C, Vertical section, highly magnified. in Devonian of the Eifel, England and North America. A, clathratum Nich. { = Stromatopora concentrica p. p., Goldf.). Clathrodidyon Nich. Like the preceding, but with radial pillars extending only between the upper and lower surfaces of successive laminae. Character- istic of Silurian ; rare in Devonian. Stromatopora Goldi. emend. Nich. (Pachystroma ^ich. and Murie) (Fig. 190). Radial pillars uniting with the thick concentric strata or latilaminae to form a finely reticulated tissue, in which tabulate zooidal tubes are sparsely distributed. Plentiful in Devon- ian ; less common in Silurian. Caunopora Lonsdale (Fig. 191), and Diapora Bargat., are Stromatoporoids which are indis- tinguishable from other genera except by the presence of numerous definitely walled tubes pene- trating the coenosteum at closer or remoter in- tervals. The tubes are often thick-walled, are furnished with horizontal or funnel-shaped tabulae, and occasionally with septal spines ; in many cases they evidently represent the corallites of Aulopora and Syringopora colonies, which have become en- veloped, but have continued to live commensally within the tissues of the Stromatoporoid. In other cases, however, the tubes appear to have been formed by true Stromatoporoid polyps. Devonian. Hermatostroma Nich. (Fig. 192). Massive or foliaceous skeletons, composed of thick parallel latilaminae, connected by vertical pillars ; pillars often running continuously through several concentric laminae. Roth pillars and laminae exhibit a dark median line when viewed in cross-section, indicating either the presence of axial canals or composition out of two lamellae. Devonian. Fio. 192. Hermatostroma sp. ind. Devonian ; Torquay, Devonshire. a, Hori- zontal lamina composed of two slightly separated lamellae ; b, Interlaminar chamberlet ; c, Radial pillar traversed by axial canal. 124 COELENTERATA— HYDROZOA phylüm ii Idiostroma Winch. Coenosteum cylindrical or fasciculate, traversed by axial, tabulate zoöidal tubes, which give off secondary lateral tubes. General tissue reticulated, similar to Stromatopora. Devonian. Labechia E. and H. Ordovician and Silurian ; North America and Europe. Stylodidyon, Stromatoporella and Syringostroma Nich. ; Amphipora Schulze ; Stachyodes Bargat. Devonian of Europe and North America. A number of genera are described by Waagen and Wentzel from the Permo-Carboniferous rocks of Farther India, such as Carterina, Disjedopora, Circopora, etc. Probably in the same neighbourhood should be placed several peculiar encrusting marine forms from the Carboniferous of Belgium, described by Gürich under the names of Aphrostroma, Spongiostroma, Chondrostroma, Malacostroma, etc. The first-named of these occurs also in the Silurian of Gotland, and was associated by Gürich with the Foraminifera. Order 4. CAMPANULARIAB AUman. (Leptomedusae, Calyptohlastea Allman ; Thecaphora Hincks). Delicate, branching, plant-like, sessile coloiiies, with chitinous periderm enveloping the base, pedunde, and also the cup-like receptades (hydrothecae) which endose the individual polyps. The proliferous zooids are developed within urn-shaped capsules (gonothecae) of comparatively large size, and sometimes become separated off as free- swimming velate Medusae. Recent Campanularians, such as are comprised by the families Sertularidae, Plumularidae and Campanularidae possess durable hard parts, but nevertheless their remains have not as yet been found in the fossil State, with the exception of a few forms from Pleistocene deposits. Range and Distribution of the Hydromedusae. Of those members of this group in which the preservation of structural parts is at all possible, the Hydrocorallinae have been recognised with certainty as early as the Upper Cretaceous. During the Tertiary they became more widely distributed, and at the present day are important reef- builders. During the Upper Jura, and notably in Tithonian beds of the Mediter- ranean region, certain genera of the Hydractinidae {Ellipsadinia, Sphaeradinia) are abundantly represented. Contrariwise, other Tubularians, such as the Triassic Heterastridium, and Parkeria and Porosphaera from the Cretaceous of central Europe, occur only sparsely. The extinct organisms known as Stromatoporoids were extremely important rock-builders during the Paleozoic, much of the limestone of the Silurian and Devonian Systems resulting from the destruction of the reefs built by these fossils. Their massive Stocks sometimes attain gigantic size. Stromatoporoid remains are profusely distributed in Ordovician and Silurian rocks of North America, England and Russia, also in the Middle Devonian of the Eifel and Ardennes, and in equivalent strata of Nassau, Devonshire, the Urals, Spain, etc. Except for a few rare survivors, the group does not Continus beyond the Paleozoic era. CLASs II GKAPTOLITOIDEA 125 ■ Appendix to the Hydromedusae. Olass or Subclass. GRAPTOLITOIDEA Lapworth.i {Rhabdophom Allman.) Under the term of Graptolitoidea are included organisms which have been considered by various authors as plant remains, horny sponges, Pennatulidae, Cephalopods and Bryozoans. Portlock, in 1843, first pointed out their analogy with the Sertularians and Plumularians ; and his inferences as to their genetic relationship were afterwards confirmed by the painstaking researches of Allman, Hall, Hopkinson, Lapworth, Nicholson, and others. More recently, however, their kinship has been denied by Neumayr and Wiman, who, on account of the bilateral symmetry of the sicula and thecae, claim that Graptolites cannot be included within any of the now existing classes of organisms. Graptolites are generally found in an imperfect state of preservation, lying flattened in the same plane upon the slaty laminae in which they are embedded, and associated in large numbers. More rarely they occur in limestone, when the internal cavities are filled with calcareous matter, and the original form accurately preserved. Such specimens have been successfully etched out and investigated under the microscope by Holm and Wiman. The general skeletal tissue (periderm) was obviously flexible, and composed of ^ Literature : Hall, J., Palaeontology of New York, vols. i., iii., 1847, 1859. — Graptolites of the Quebec Group. Cauad. Organic Remains, dec. ii. Geol. Surv. Cauad., 1865. — lutroduction to the study of Graptolites. 20th Ann. Rept. N.Y. State Gab. Nat. Hist., 1868.— Barrande, J., Grapto- lites de Boheme. Prague, 1850. — 8uess E., über böhmische Graptolithen. Haidinger's Naturw. Abhandl., 1851, vol. iv. — Scharenherg, PF., Ueber Graptolithen. Breslau, 1851. — Geinitz, H. B., Die Versteinerungen der Grauwackenformation in Sachsen, Leipzic, 1852. — Die Graptolithen des mineral. Museums in Dresden, 1890. — Richter, R., Thüringische Graptolithen. Zeitschr. Deutsch. Geol, Ges., vols. v., xviii., xxiii., 1853, '66, 71. — Nicholson, II. A., Monograph of the British Graptolitidae, 1872. — Lapworth, C, Notes on the British Graptolites. Geol. Mag., vols. x., xiii., 1873, '76. Also various articles in Quart. Journ. Geol. Soc, 1875, '78, '81, and Ann. Mag. Nat. Hist., 1879, '80.— On the Graptolites of County Down. Ann. Rep. Belfast Nat. Field Club, 1877, vol. i., pt, iv, — Tullberg, S. A,, On Species of Didymograptus. Geol. För. Stockholm Förh,, 1880, vol. v. — Spencer, J. W., Graptolites of the Upper Silurian System. Bull. Mus. Univ. Missouri, 1884. — Tornquist, S. L., Observations on Graptolites. Acta Univ. Lund, 1890-92, vols. xxvii.-xxix. — Holm, G., Skandinaviens Gi'aptoliter. Svensk, Vetensk. Akad. Förh., 1881, vol. xxxviii. — Gotland's Graptoliter. Bihang Svensk. Vetensk, Akad. Handl. 1890, vol. xvi. — Barrois, C, Memoire sur la distribution des graj)tolites en France. Annales Soc. Geol. Nord, 1892, vol. xx.— Wiman, C, Über Monograptus und Diplograptidae. Bull. Geol. Inst. Upsala, 1893, vol. i. (English translatiou in Journ. Geol., 1893, vol. ii.). — Tornquist, S. L., Observations on the Structure of some Diprionidae. Fisiogr. Sällsk. Handl., 1893, '97, vols. iv., viii. — Researches into the Graptolites of the Scanian and Vestrogothian Phyllo-Tetragraptus Beds. Lunds Univ. Arsskrift, 1901, vol. xxvii. — Ferner, J., liltudes sur les graptolites de Boheme. Prague, 1894-99. — Holm, G., Om Didymograptus, Tetra- graptus och Phyllograptus. Geol. För. Förh., 1895, vol. xvii. No. 164. — Ruedemann, R., Synopsis of the Mode of Growth and Development of the Genus Diplograptus. Amer. Journ. Sei. (3), 1895, vol. xlix. ; Also in Ann. Rept. N.Y. State Geol. 1894, and Amer. Nat., 1897, vol. xxxii. — Graptolites of New York, N.Y. State Museum. Memoirs vii. 1904, and xi., 1908. — Gurley, R. R., North American Graptolites. Journ. Geol., 1896, vol. iv. — Wiman, C, Über die Graptolithen. Bull. Geol. Inst. Up.sala, 1895, vol. ii. ; also Nat. Sei., 1896, vol. ix., and Bull. Geol. Inst. Upsala, 1897, No. 6. 1900, No. 10. — EUes, G. L., and Wood, E. M. R., Monograph of British Graptolites, ed. by C. Lapworth. Palaeontogr. Soc, 1901 to date. — Hall,T. S., Note on the Distribution of the Graptolitliidae in the Rocks of Castlemaine. Rept. Austral. A.ssn. Adv. Sei., 1894 ; also Proc. Roy. Soc. Victoria 1892, '97, '98, '99 ; Geol. Mag,, 1899, vol. vi.—Roemer, F., and Frech, F., Lethaea Palaeozoica, 1897, vol. i. — Bassler, R. S., Dendroid Graptolites of the Niagarau Dolomites at Hamilton, Ontario. Bull. Smithson. Inst., No. 65, 1909. — Westergard, A. H., Studier öfver Dictyograptus.skiffern. Med. fr. Lunds Geol. Fält Klub. Ser. B, No. 4, 1909. 126 COELENTERATA— HYDROZOA phylum ii smooth or fiiiely striated chitine ; usually it has the form of a dense continuous membrane, but in the Retiolitidae it is attenuated and supported by a latticed network of chitinous threads. It is usually preserved as a thin bitumino- carbonaceous film, which, however, is often infiltrated with pyrites, and is not infrequently replaced by a glistening greenish-white Silicate (Gümbelite). The Compound organism or rhabdosome ('^ polypary ") of the Graptolites is usually linear, more rarely petaloid in form, undivided or branching, and is either straight, beut, or in exceptional instances spirally enrolled. These rhabdosomes, each of which originates from a sicula (see below) may again be united into colonies of a higher order (synrhabdosome). Cup-shaped rhabdo- thecae, which are usually obliquely set and more or less overlapping, are borne on one or on both sides of the polypary, and are united by a common coenosarcal canal enclosed in the /periderm. The polypary is in later forms strengthened by a peculiar chitinous axis (virgula, solid axis), which in the Monograptidae runs in a groove lying outside the coenosark on the dorsal side of the organism {i.e. on the side opposite to the theciferous margin). But in the biserial Graptolites the virgula is either enclosed between the laminae of a central or sub-central septum, which is formed by the coalescence of the flattened dorsal walls (Biprionidae) ; or it is double and the two virgula (see text Fig. 209) are placed on opposite sides of the coenosark, and are united with the peridermal network {Retiolitidae). Springing from the common canal, is a series of thecae {celhles, denticles), which are disposed in longitudinal rows along either one (Fig. 193), two (Fig. 194) or four sides of the polypary. They usually have the form of elongated, cylindrical, rectangular or conical sacs; their walls are in most cases applied to those of their neighbours above and below, although occasionally they spring out quite isolated from one another. Each theca opens directly into the common canal, and is furnished distally with an external aperture, the form and size of which vary extremely in difFerent species. In some forms it is circular or quadrate or introverted or introtorted ; in others it is contracted. Not infrequently the outer lip is ornamented with one or two slender spines, which often subdivide and inosculate with one another. The form of the thecae and apertures has been employed by Lapworth to define families and subfamilies. The polypary in most Graptolites is furnished at its proximal end with a minute, triangulär or dagger-shaped, originally conical, body called the sicula (Fig. 195), which represents the original embryonic skeleton and is suspended from an originally tubulär filament, the 7iema oy nemacaulus {Fig. 196). In the wall of the sicula is formed, in the later Graptolites, an axis or rod, the virgula, which extends through the rhabdosome. Rhabdothecae are then budded either uniserially along one side, or in alternate sequence along both of the lateral margins of the sicula, originating from one theca near the major end of the sicula. They grow either laterally away from the sicula (Axonolipa) or along the nemacaulus (Axonophora). The sicula itself ceases to grow, as a rule, after the first thecae are budded, and sometimes it becomes obsolete or absorbed. Sometim^es the rhabdosome remains undivided, sometimes it forms branches, which may diverge at various angles ; in other cases two or four uniserial polyparies may be placed back to back with their dorsal walls coalescing, thus giving rise to di- or tetra-serial colonies. In the latter types the coenosarc is commonly divided by one or two median septa. I GRAPTOLITOIDEA 127 Graptolites commonly occur in argillaceous schists, moie rarely in limestone formations, of the Upper Cambrian, Ordovician and Silurian Systems. They seem to have swarmed in the muddier portions of the sea, and floated either attached to sea-weeds, etc., or as free-swimming colonies ; or, in rarer instances, remained stationary with the sicula or a root embedded in the mud, or attached to foreign bodies. They are divided into two Orders : Dendroidea Nicholson {Cladophom Hopkinson), and Graptoloidea Lapworth, or Graptolites proper. The :^"^?- Fig. 195. a, Münograptus gregarius Lapw. Sil- urian ; Dobbs, Linn, Scotland. Proxiinal end showing sicula, enlarged ; b, Didy- mograptus pennatulus Hall. Ordovician (Quebec Group) ; Point L6vis, Canada. Proximal end .showing sicula, enlarged (after Lapworth). Fig. 193. A, C, Monograptus priodon (Bronn). Silurian (Etage E); Prague. A, Rhabdosoma, natural size. B, Longitudinal section, enlarged. C, Dorsal aspect, enlarged. D, Mono- graptus hnhemicus Barr. Same locality. a, Virgula ; c, Com- mon canal ; th, Thecae ; x, External aperture (after Barrande). Fig. 194. a, c, Climacograptus typicalis Hall, Ordovician (Trenton limestone) ; Cincinnati, Ohio. a, Vertical section, enlarged, showing central virgula ; b, Individnal of the natural size ; c, Cross-section, enlarged ; d, e, Diplograptus palmeus Barr. Silurian ; Prague. Rhabdo- soma of the natural size and several times enlarged (after Barrande) ; /, Diplograptus foliaceus Murch, Ordovician (Llandeilo Group); Scotland. natural size (after Lapworth). Fig. 196. Dictyonema cavernosum Wiman. Or- dovician ; Gotland. Proximal end of rhabdosome with adhesion disk (x), large nourishing individnal (z), and sniall budding individnal or gonangium (zi). 3/i (after Winian). latter are again divided into two suborders : AxonoUpa Frech, redefined by Ruedemann, without axis or virgula; and Jxonophai-a Frech, limited by Ruedemann, with an axis. ^ Order 1. DENDROIDEA Nicholson. Family. Dendrograptidae Roemer. This family, which includes all dendroid forms, is represented during the older Paleozoic by the genera Deiidrograptus, Dictyonema, Desmograptus, Gallo- 128 COELENTERATA— HYDROZOA PHYLUM II gmptus and Ftilograptus. Their polyparies are finely branching and plant-like in appearance, sometimes furnished with a strong foot-stalk, in other cases terminating acutely at the base. The original substance was undoubtedly chitinous. In well-preserved specimens are seen on one or occasionally on both sides of the branchlets numerous small cellules or thecae, in which evidently the zooids were seated. These thecae have been shown by Wiman to be of threefold character, some of them having lodged nourishing, others budding and others sexual individuals or gonangia. Very often the branches of the dendroid rhabdosome are united by numerous delicate processes or Dendrograptus Hall. Rhabdosome consisting of a strong main stem and a broad, spreading, shrublike, variously ramifying frond. Thecae commonly Fig. 197. Didyonema fläbelliforme (Eichw.). Upper Cambrian; Rensselaer County, N.Y. A, Sicula with very long neina, x =Vi- -ß» Mature rhabdosome with adhesion stem, x ^/i (after Ruedemann). obscure, but sometimes distinct and angular, or they may occur as round or elliptic pits or pustules, Cambrian to Silurian. Ftilograptus Hall. Rhabdosome with branches giving ofF branchlets alternately on opposite sides, the general habit being suggestive of Recent hydrozoans. Ordovician and Silurian ; eastern North America. Didyonema Hall {Didyograptus Hopkinson) (Figs. 196, 197). Ehabdosomes forming funnel or fan-shaped fronds, composed of numerous bifurcating branches arising from an acute base, and united at intervals by fine dissepi- ments. Thecae with complicated appendages, their branches supporting three kinds of individuals, nourishing, budding and sexual. Cambrian to Car- boniferous ; especially abundant in Ordovician of Norway, but usually com- pressed into a basket-like network. Desmograptus Hopkin. Differs from the preceding in the flexuous char- acter of the branches, which coalesce at intervals ; dissepiments chiefly in lower part of the frond. Ordovician to Devonian ; Europe and North America. CLASS II HYDEOZOA— GRAPTOLITOIDEA 129 Order 2. GRAPTOLOIDEA Lapworth. Suborder A. AXONOLIPA Frech (emend. Ruedemann). Family 1. Dichograptidae Lapworth. Uniserial Graptolitoidea tvith bilateral rhabdosome ; brauch es dichotomous ; thecae simple^ sub-cylindrical. Dichograptus Salter (Fig. 198). Rhabdosome consisting of eight simple Fia. 199. Tetragraptus hryonoides Hall. Ordovician ; Point L6vis, Canada (after Hall). Fig. 200 Didymograptus pennatulus Hall. Ordovician ; Point L^vis, Canada (after Hall). Fig. 198. Dichograptus octobracMatus Hall. Ordovician (Quebec Group) ; Point L^vis, Canada (after Hall). Fio. 202. Didyynograp' l'hyllograjnus typv,n Hall. t)is murchisoni Ordovician (Quebec Group); Point L^vis, Canada. a, Several polyparies of the natural size ; b, Ideal cross-section, enlarged (after HaH). (Beck). Middle Ordovician (L 1 a n d e i 1 0 Group) ; Wales. uniserial branches which are produced by repeated dichotomy, and their bases often enveloped in a central corneous disk. Ordovician. Tetragraptus Salter (Fig. 199). Rhabdosome consisting of four uniserial branches which are produced by twice repeated dichotomous division. Ordovician. Didymograptus M'Coy (Figs. 195, J; 200; 201). Rhabdosome consisting of two symmetrical branches diverging from a small primary cell (sicula) at various angles. Thecae obliquely directed, having the form of flattened rectangular prisms, and in contact with one another throughout. Ordovician. Phyllograptus Hall (Fig. 202). Rhabdosome leaf-like, composed of four VOL. I K 130 COELENTERATA phylum ii uniserial rows of prismatic thecae coalescing along the whole length of their dorsal margins ; the entire structure resembling Tetragraptus with the four branches grown together, each two back to back and forming a cross in transverse section. Ordovician. Family 2. Leptograptidae Lapworth. Uniserial, flexuous, bilateral rhabdosomes, with simple or Compound lateral branches; thecae with a slightly sigmoid curvature, apertures inclined, somewhat introverted. Leptograptus Lapworth. Rhabdosome consist- ing of two long, filiform, bilaterally symmetrical branches. Ordovician. Nemagraptus Emmons {Coenograptus Hall) (Fig. 203). Two primary branches originating from the Fir.. 203. centre of a triangulär sicula, more or less flexed, Coenograptus iiraciiis Hall. Orrio- and giving off simplc branches from the convex cian- ""•■"'' ^"■"- ''" '■■ '■'""•- CO Hall). vician : Point Levis, Ciinada (after •!. '.i i-^ ^ r\ ^ ' • sideatapproximately regulär intervals. Ordovician Family 3. Dicellograptidae 'Lapworth. Uniserial or uni-biserial Graptoloidea. Thecae tubulär, with conspicuous sigmoid ventral curvature. Apertures situated in excavations and frequently intro- verted and introtorted. Dicellograptus Hopk. Rhabdosome bilaterally symmetrical, consisting of two uniserial branches diverging from the sicula at angles ex- ceeding 180 degrees. Ordovician. Dicranograptus Hall (Fig. 204). Rhabdosome Y-shaped, com- posed of two symmetrically developed branches which are coalescent in the proximal and free in the distal portion of their length. Ordovician. Suborder B. AXONOPHORA Frech (emend. Ruedemann). Family 1. Diplograptidae Lapworth. Biserial Graptolitoidea with rectilinear rhabdosomes. pjo 204. Climacograptus Hall (Fig. 205, a, c). Rhabdosome bilaterally ramZT^Tl^i. symmetrical. Thecae tubulär, ventral walls sigmoidally curved, ^Hudson mip^v apertural margin horizontal, situated within a well-defined ex- New York (after cavation. Ordovician and Silurian. Diplograptus M'Coy (Figs. 205, d-f ; 206). Rhabdosome bilaterally sym- metrical, rectangular, concavo-convex or tabular in section. Thecae mostly sub-prismatic, ventral walls inclined and straight. Subgenera : Orthograptus and Glyptograptus Lapworth ; Mesograptus Elles and Wood ; Petalograptus Suess ; Cephalograptus Hopkinson. Ordovician and Silurian. Family 2. Glossograptidae Lapworth. Biserial Graptoloidea with straight rhabdosomes, fest attenuated, with framework of strengthening ßbres. Thecae of Diplograptid type, provided with spurs and other processes which often form an external lacework. Glossograptus Emmons (Figs. 207, 208). Rhabdosome having lingulate CLAss n HYDROZOA— GRAPTOLITOIDEA 131 outline and rounded extremities, ornamented with two rows of isolated spurs ; Fig. 205. a, c, Cliiruicograptus typicalis Hall. Ordovician (Tren- ton limestone) ; Cincinnati, Ohio ; a, Vertical section, showing common axis in the centro, enlarged ; 5, Poly- paiy of the natural size ; c, Cross-section, enlarged. d, e, Di'plo(}raptus jmlmeus Barr. Silurian ; Prague ; d, Polypary of the natural size ; e, Polypary enlarged. /, Diplograftus foliaceus Murch. Silurian (Llandeilo Group) ; Scotland. Natural size. Fio. 206. Diplograptus foliaceus Hall.iUtica shale (Ordovician); Dolgeville, New York. Synrhabdosome showing central organs and primary disk (hl) with funiculus (/), to which the rhabdosomes (r) are attached by a slender nemacaulus (n). Gonangia (f/) with young siculae (s) are also present. x % (after Ruedeniann). Fkj. 207. Glofisogriiptus (luadrimiicronatus Hall var. approximatus Rued. Utica shale (Ordovician) ; Dolgeville, New York. Synrhabdosome. x Vi (after Ruedeinann). Fio. 20S. Glossograptus quadrimucronatus (Hall) var. apjyroximatus Rued. Utica shale (Ordovician) ; Dolgeville, New York. Young synrhabdosome showing central disk and siculae. x »/i (after Ruedeniann). each theca with two long spines. Ordovician. lletiograptus Hall ; Lasiograptus Lapworth. Ordovician. 132 COELENTERATA PHYLUM II Family 3. Retiolitidae Lapworth. Biserial Graptolitoidea with straight rhahdosomes, the latter characterised hy a neiwork of delicate chitinous tracery (reticula) which forms the outward covering of the walls of the thecae. RetioUtes Barr. (Fig. 209). Rhabdosome with periderm attenuated and supported on a meshwork of fibres. Thecae arranged biserially, their apertures opening outward. Two virgulae attached to opposite sides, in the median plane. Ordovician and Silurian. Subgenera : Gladiograptus Hop- kinson and Lapworth ; Gothograptus Frech. Family 4. Dimorphograptidae Lapworth. Uni-biserial Graptolitoidea, in which the proximal portion is uniserial, hearing thecae of the general Monograptus type ; the distal portion is biserial with thecae of the Diplograptid type. Dimorphograptus Lapworth. Silurian. Family 5. Monograptidae Lapworth. Uniserial Graptolitoidea, with simple or Compound, straight or convex rhabdosome and thecae of varied form. ^^^- 20^- Monograptus Geinitz {Monoprion Barrande ; RetioiitesgeinUzianusB.vv. mun^n^ Pomatograptus and Pristiograptus Jaekel) (Figs. 193, 210). Rhabdosomes with only a single row of thecae, which are in contact, usually overlapping, their apertures entire or contracted, often directed downward. Form of the rhab- A, Specimen from siliceous Feuguerolles, Calvados ; natural size B, C, Polyparies from Motala, Sweden. B, Cross-section. C, Lower end, en- larged ; calcareous matter dissolved out by acid. v, Zigzag-shaped virgula ; ■y', Rod-like virgula ; th, Conjoined walls of hydrothecae ; s', Crossbars connecting the virgulae ; o, Apertures dosome may be Straight, curved or somctimes spirally coiled. Silurian and Devonian. Bastrites Barr. (Fig. 211). Rhabdosome simple, spirally coiled; common canal very narrow ; distal parts of thecae more or less linear and widely separated from one another. Silurian. The zonal distribution of species in Thuringia and Saxony has been worked out in detail by Eisel. Cyrtograptus Carruthers. Variously branching Monograptidae. Silurian. Range and Distribution of Graptolites. Graptolites are excellent index fossils of the older Paleozoic rocks, owing to their limited vertical ränge, and wide geographical distribution. The simpler forms, such as are derived by a succession of budding from a primary sicula (Axonolipa), are especially characteristic of the uppermost Cambrian and lower half of the Ordovician rocks. The group as a whole becomes extinct at the close of the Silurian, except for a few stragglers in the Devonian and Carboniferous. The occurrence of these organisms in rocks of the same age in all parts of the world is explained by the fact that while some forms were w SUBCLASS II HYDROZOA— ACALEPHAE 133 probably attached to seaweeds, as often in modern hydroids, others were free- floating or planktonic creatnres. Remains of Graptolites are profusely distributed in the siliceous schists and alum slates of the Fichtelgebirge, Thuringia, Saxony and Bohemia. They are plentiful also in the Harz, in Poland, Silesia, the Baltic Provinces and the Ural district ; and again in Scandinavia, Cumberland, Wales, the north of England, Scotland and Ireland, as well as in Normandy, Brittany, Spain, Portugal, Sardinia and Carinthia. In America they are found ex- a ' h c quisitely preserved in Newfoundland, Canada, New York, Virginia, Alabama, Ohio, Wisconsin, Iowa and Arkansas. They Fig. 210. a, Monograptus nilssoni Barr. Silurian (Alum Schists) ; Gräfenwerth, near Schleitz, Germany; h, Monograptus colonus Barr. Silurian ; Eliots- tieid, Scotland, showing sicula (after Lap- worth) ; c, Monograptus turriculatus Barr. Sil- urian ; Prague (after Barrande). All figures natural size. Fig. 211. Rastrites Knnaei Barr. Silurian ; Zekko- witz, near Prague (after Barrande). are known also in South America (Bolivia), and Australia, and are not un- common in the drift which Covers the plains of Northern Germany. According to Lapworth, Graptolites are distributed vertically throughout six difFerent horizons ; 'the first of these coincides with the Upper Cambrian, the three foUowing with the Ordovician, and the two uppermost with the Silurian. The Monoprionidae are especially characteristic of the two Silurian horizons. [The discussion of tlie group Graptolitoidea in the present work has been revised by Dr. Rudolf Riiedemann, of the New York State Geological Survey, at Albany, New York. — Editor. 1 Subclass 2. ACALEPHAE Cuvier. Scyphomedusae.^ {Discophora Huxley). Free-swimming, discoidal or bell-shaped Medusae, with downwardly direded mouth, with gastro-vascular pouches and numerous radial canalSj and having, as a rule, the margin of the umhrella lobed. Cambrian to Eecent. The Acalephs or Lobed Jelly-fishes, though frequently of considerable size, are entirely without hard parts, and therefore are unfitted for preservation ^ Literature : Huxley, T., Memoir on the Anatomy and Affinities of the Medusae. Phil. Trans,, 1849. — K7ier, R., Ueber eine Meduse in Feuerstein. Sitzungsber. Akad. Wiss. Wien, 1865, vol. lii. — Haeckel, E., Ueber fossile Medusen. Zeitschr. für Wissenschaft. Zool., 1865 and 1870, vols. xv., xvii. Neues Jahrb. für Mineral., 1866. Jenaische Zeitschr., 1874, vol. viii. System der Medusen, i. and ii., Jena, \^^0-^\.—Nathorst, A. G., Om Aftryck af Medusor, etc. K. Svenska Vetensk. Akad. 134 COELENTERATA PHYLUM II in the fossil state. Nevertheless, under exceptionally favourable conditions, as, for instance, in the Upper Jurassic Lithographie Stone of Bavaria, and in the Middle Cambrian shales of British Columbia, impressions of these delicate organisms are sometimes preserved, which admit of precise determination. The best preserved and at the same time the most abundant species is Rhizostomites admirandus Haeckel, belonging to the Acraspedote family of Rhizostomidae (Fig. 212). Impressions also occur in flinty concretions of the Upper Cretaceous, which are most nearly referable to the Medusae. Of a more questionable nature are the organisms occurring in the Cambrian sandstone of Lugnaes, Sweden, described by Thorrell under the name of Rhizostomites admirandus Haeck. Lithographie stone ; Eichstädt, Bavaria. 1/7 natural size. (Missing parts restored'in outline.) Spatangopsis, but assigned by Nathorst to the Acalephs. In the same strata also are found those peculiar fucoidal structures known as Eophyton, which are commonly supposed to be of vegetable origin. Nathorst has brought forward evidence, however, to show that these may really have been produced by the trails of Jelly-fishes. Here also should be noticed the forms described by Nathorst as Medusites, frona the Lower Cambrian of Sweden, and regarded by this author as casts of the gastric cavity of Jelly-fishes. In 1898 a valuable monograph on fossil Medusae was contributed by Walcott, and in 1911 our knowledge of these organisms was increased in important respects by the same author, as a result of his studies of remarkably well-preserved specimens from the Cambrian of British Columbia. Handl., 1881, vol. xix.—Ammon, L. v., Ueber jurassische Medusen. Abhandl. Bay. Akad. 1883, vol. xvii. — Brandt, A., Ueber fossile Medusen. Mem. Acad. Imp. St-Petersb., 1871, 7tli ser., vol. xvi. — Pohlig, H., Altpermische Medusen. Festschrift zum 70ten Geburtstage K. Leuckarts, U^2.- -Walcott, ö. D., Fossil Medusae. U. S. Geol. Surv. Monogr., xxx., 1898.— /c^em, Middle Cambrian Holothurians and Medusae. Smithson. Mise. Coli., 1911, vol. lix. No. 3. — Mayer, A. 0\, The Medusae of the World, i.-iii. Carnegie Inst. Wash,, Pub. No. 119, 1911. Phylum III. VERMES. Worms.' Bilaterally symmetrical animals with unsegmented or uniformly segmented, and usually elongated bodies having a distinct body cavity. Segmented lateral appendages wanting. A dermal muscular sysj^em and paired excretory canals (water-vascular System) present. Of all the larger divisions of the animal kingdom, none is so poorly adapted for preservation in the fossil State as the Worms, whose bodies are as a rule entirely destitute of hard parts. All Worms are bilaterally symmetrical, and dorsal and ventral surfaces are clearlj^ differentiated. The unsegmented Worms (Vermes proper) have either fiat or cylindrical bodies, and are accordingly distinguished as Platyhelminthes or Fiat Worms, and Nemathelminthes or Round Worms. But with the exception of the Cambrian genus Amiskwia (Fig. 213), supposed to be allied to the Recent Sagltta, and a few rare parasitic forms discovered in Carboniferous insects, or in Tertiary insects enclosed in amber, neither of these classes is represented in the fossil State. • The segmented Worms, or Annelida, are characterised by a division of the body into metameres, which, although primitively alike, do not always remain homonomous. They have a brain, a circumoesophageal ring, a ventral chain of ganglia, and a vascular System. The body is more or less elongated, and ^ Literature : Fander, O. H., Monographie der fossilen Fische des silurischen Systems des russisch-baltischen Gouvernements, 1851. — Ehlers, A'., Die Borstenwürmer (Annelida Chaetopoda). Leipzic, 1864-68. — Idem, Über fossile Würmer aus dem lithographischen Schiefer in Bayern. Palaeontogr. , 1868, vol. xvii. — Olaparede, Vi'., Recherches sur la structure des Annelitles sedentaires, 1873. — Newherry, J. *S'., Palaeontology of Ohio, vol. ii. part 2, 1875. — Ilinde, G. J.. On Conodonts from the Chazy and Cinciunati Groups ; and on Annelid Jaws from the Cambro-Silurian, Silurian, and Devonian Formations in Canada, and from the Lower Carboniferous in Scotland. Quar. Journ. Geol. Soc, 1879, vol. xxxv. — Ulrich, E. 0., Journ. Cinciunati Soc. Nat. Hist., 1879, vol. i. — Hitide, G. J., On Annelid Jaws from the Wenlock and Ludlow Formations of the West of England. Quar, Journ. Geol. Soc, 1880, vol. xxxvi. — Etheridge, R.,jun., British Carboniferous Tubicolar Annelida. Geol. Mag., 1880, vol. vii. — Nathorst, A. G., On the Tracks of sonie Invertebrate Animals and their Palaeontological Significance. K. Svensk. Vetensk. Akad. Handl., 1881-86, vols. xviii., xxi. — Ilinde, G. J., On Annelid Remains from tlie Silurian Strata of the Isle of Gotland. Bihang tili K. Svensk. Vetensk. Akad. Handl., 1882, vol. \n.~Zittel, K. A., and Rohon, J. V., Ueber Conodonten. Sitzber. Bay. Akad. Wiss., 1886, vol. xvi.—Clarke, J. M., Annelid Teeth from the Lower Portion of the Hamilton Group, New York. Sixth Annual Report, N.Y. State Geologist, 1886. — Rovereto, G., Studi monografici sugli Annelidi fossili. Palaeont. Ital., 1904, vol. X.— Walcott, O. D. , Middle Cambrian Annelids. Smithson. Mise. Coli. ,1911, vol. Ivii. No. 5. 135 Fig. 213. AmislauiM sayltti- formis Walcott. Mid- dle Carnbrian; British Columbia. Flatt«ned specimen, x ^/i (after Walcott). 136 VEEMES PHYLUM III sometimes flattened, sometimes cylindrical. According as the internal Segments correspond exactly with the external, or as each internal segment corresponds to a definite number (3, 4 or 5) of the external rings, two classes, Chaetopoda and Hirudinea, are distinguished. A further difference is to be noticed in the locomotive organs, the Chaetopoda having bristle-bearing, unjointed appendages (parapodia) on each ring of the body ; and the Hirudinea having a terminal sucker. The latter group includes only the Leeches, which are not known with certainty in the fossil State. Fossil representatives of the third class, the Gephyrea, Annelids with the body devoid of any appearance of segmentation in the adult condition, are known ; but of the fourth and last class Archiannelida, the most primitive of all living Annelids, no fossil remains have been found. Class 1. CHAETOPODA. (Earthworms, Annelids, etc.) It is only with the subclass of mari>ne worms (Polychaeta) that the paleontologist is concerned since the earthworms and their allies (Oligochaeta) are wholly unknown as fossils. The marine Chaetopoda are divisible into three Orders, the Miskoa, the Tubicola or Sedentary Worms, and the Nereid or Errant Annelids. Order 1. MISKOA Walcott. Polychaeta with similar segments and parapodia throughout the length of the hody ; retradile prohoscis ; straight enteric canal. Body not distinctly speciaUsed into sections. This Order is founded upon a remarkable series of Annelids discovered by HP^r'^rlJ , Wl ffjgKKl |BfcL, V- i^k 1^^ ^1 Fig. 214. Wiwaxia corrvgata Walcott. Middle Cambrian ; British Columbia. Crushed specimen showing displaceü spines and scales, X i/i (after Walcott). i'Kis. zia aiid 216. Cambrian Polychaeta from British Columbia (after Walcott). Canadia spinosa and Aysheaia ■pedunculata Walcott, both x ^/a- Walcott in the Middle Cambrian of British Columbia and described by him in 1911. The Order is represented by four families, namely Miskoidae and Aysheaidae, with the genera Mislwia and Aysheaia respectively ; Canadidae including Canadia and Selkirkia ; and the Wiwaxidae with the three genera LASS I CHAETOPODA 137 'iwaxia, Pollingeria and JVorthenella. Typical examples of these Annelids are shown in Figs. 214-16. Frotoscolex and Eotrophonia Ulrich, from the Eden Fig. 217. Gephyrean Annelid, I'ikaia gracücns Walcott. Middlc Cambrian ; British Columbia, x 2/j (from Walcott). shale of the Ohio Valley, are probably Ordovician representatives of this Order. Order 2. TUBICOLA. (Sedentaria.) Pohjchaetous Annelids with indistindly separated head, and short, usually non- protrusible prohoscis, witJiout jaws. Parapodia short, and never used for swirnming . Inhabiting more or less firm tuhes, which they construd, and subsisting upon vegetable matter. The Tubicolous Annelids invest themselves with a protective tube of more or less irregulär form, to which they are not organically attached, and within which they can move freely. Sometimes the tubes are free, but more com- monly they are attached to foreign objects, either by the apex or by one side, and may occur either singly or in Clusters. The tubes frequently consist of concentric layers of lime-carbonate, with vesicular cavities between the lamellae, or the latter may be traversed by fine tubuli. In other cases the tubes are composed of agglutinated grains of sand and other foreign particles ; or they may be membranaceous or leathery. The materials for constructing the tubes are procured by the tentacles or branchial filaments of the head, and are cemented together by a glutinous secretion from large glands. Fossil worm-tubes are by no means of infrequent occurrence, and are known from the Ordovician onwards. Only a few of the more common examples can be mentioned here. Serpula Linn. (Fig. 218). ünder this head are included the majority of fossil Tubicolous Annelids. They build firm, irregularly contorted, sometimes spirally enrolled, free or adherent calcareous tubes, which are frequently clustered together in large numbers. Beginning in the Silurian, they are sparsely represented in the Paleozoic era ; but from the Jura onward, nunier- ous forms occur, the usual condition being attached upon other fossils. Notably in the Lower Cretaceous their gregarious masses form beds of con- siderable thickness (Serpulitenkalk of Brunswick, and Serpulitensand of Banne witz, near Dresden). . S. spirulaea Lam. (Fig. 218, H) is an abundant and characteristic Eocene species. Eecent Serpulas have a world - wide distribution. 138 VERMES phylüm iii Terehella Cuv. (Fig. 218, I). Cylindrical, elongate, more or less beut Fi(!. 218. A, Serpula Umax Goldf. Middle Jura ; Franconia. ß, C, S. gurdialis Schloth. Upper Cretaceous : Banne- witz, near Dresden. D, S. convoluta, Goldf. Middle Jura; Stuifen, Wurtembercr. E, S. socialis Goldf. Middle Jura ; Lahr, Baden. F, Same, enlarged. G, S. septevisulmta Reich. Upper Cretaceous ; Bannewitz. H, S. (Rot.ulariaJ)efr.)spirulaeaIjam. Eocene ; Monte Berici, near Vicenza. I, TerpheUalapilloides Münster. Upper Jura ; Streitberjr, Franconia. tubes, composed of cemented grains of sand, fish-debris, or other adventitious particles. Lias to Eecent. Spirorbis Daudin {Microconchus Murch.) (Fig. 219). Minute, snail-like or spirally enrolled calcareous tubes, cemented by the flattened under side. The spiral may be either dextral or sinistral, and is usually ornamented externally with concentric Striae or annulations, sometimes with tubercles or spines. Abundant in the Paleozoic forma- tions from the Ordovician onward, and also at the present day ; somewhat less common in the Mesozoic and Cenozoic eras. Kecent species usually adherent on seaweeds. Fig. 219. Spirorhis omphalodes (Goldfuss). Tubes seated upon a Brachiopod shell {Schiichertella umhraculnm). Devonian ; Gerolstein, Eifel. The foUowing genera are regarded as Annelids, but their systematic position is doubtful : commonly CLASS I CHAETOPODA 139 Serpulites Mnrdh. Very long, smooth, compressed, and somewhat bent calcareous tubes, the layers admixed with organic substance. Ordovician and Silurian. CornuUtes Schloth. Thick-walled, trumpet-shaped tubes, Serpula-like at the lower end, and sometimes attaining a length of three or four inches. Exterior annulated, and covered with very -fine longitudinal Striae. Some authors regard the tubes as Pteropod shells. Ordovician to Devonian. Ortonia Nich. Small, conical, slightly flexuous, thick-walled calcareous tubes, cemented by the whole of one surface to some foreign body. Sides of the tube ringed with imbricating annulations, the free upper surface apparently cellular in structure. Ordovician to Carboni- ferous. Conchicolites Nich. Conical, slightly bent, thin-walled tubes, growing together in Clusters, and attached by the small lower ends to orthoceratite or Brachiopod shells. Tubes made up of numerous short rings, each of which partially overlaps the subjacent one. Ordovician. The peculiar group Myzostomidae, which are external parasites on Kecent Crinoids, are thought to be related to the Chaetopoda. Graff has shown that they also infested the column Segments of Jurassic Crinoids. Order 3. ERRANTIA. (Nereidae). Free - swimming, predaceous Polychaeta, with well-marJced liead. Proboscis capoMe of protrusion, and armed with papillae or powerful jaws. Para- podia 7nuch more developed than in the Tubicola, heset with setae, and serving for locomotion. Undoubted remains of Errant Worms have long been known from the Lithographie Stone (Upper Jura) of Bavaria, and include the trails, calcified jaws and excrements of num- erous species. The principal genus from this horizon is Eunicites Ehlers {Geophilus Germar) (Fig. 220), perfect impressions of which are also found in the Upper Eocene limestone of Monte Bolca, Italy. Archarenicola Horwood is known from the English Rhaetic. Uiider the designation of Lumhricaria Münster (Lumbricites Schlotheim) (Fig. 221) are included a variety of obscure. remains from the Lithographie Srone, which may be best regarded as the excrements of Annelids. They occur as irregularly contorted bands or strings, sometimes in the form of very long labyrinthic coils. Of peculiar interest are the minute detached jaws and denticulated plates Fig. 220. Eunicites avitus Ehlers. Lithographie Stone ; Eichstädt, Bavaria. Natural size. 140 VERMES PHYLUM III described by Hinde in the Ordovician, Silurian, Devonian'and Carboniferous rocks of the United States, Canada, Great Britain and Sweden (Island of Got- land). These are very small, black, highly lustrous bodies, extremely variable in form (Fig. 222), and mainly composed of chitinous material which is unafFected by acid. They exhibit a striking resemblance to the jaws of recent. Annelids, and probably represent a large number of genera. Of less certain derivation are the microscopic teeth first described by Pander under the name of "Conodonts" (Fig. 223), which occur detached in the Cambrian (Blue Cläy underlying the Ungulite Grit) of St. Petersburg, and are also very abundant in beds of Ordovician, Silurian and Car- boniferous age in Russia, Great Britain, the United States and Canada. They are usually translucent, lustrous or corneous, and are composed Fig. 231. Lumhricaria colon Münst. Lithograi)hic Stone ; Solen- hofen, Bavaria. Natural size. -3 Fig. 222. Paleozoic Annelid-jaws. A, Lumbriconereites hasalis Hinde. Silurian ; Dundas, Ontario. lo/j. ß^ Oenonites rostratus Hinde. Toronto. l5/i. C, Eunicites varians Grinnell. Toronto, "/i- ^, Arabellites scutellatus Hinde. Ordovician ; Toronto, lö/j. of carbonate and phosphate of lime. They exhibit very great variety in form. By Pander and others these fossils have been regarded as fish-teeth. Zittel ^ Fig. 223. Conodonts, greatly enlarged. A, B, Paltodus truncatus Pander (after Pander). C, Prioniodus elegans Pander. Cambrian ; St. Petersburg. D, Polygnathus duUus Hinde. Devonian ; North Evans, NeAv York. 20/^. and Rohon, however, consider that they are Annelid jaws, but their true Position cannot yet be said to have been positively determined. CLASS II GEPHYREA 141 Fig. 224. Nereites camhrensis M'Leay. Cambrian ; Llampeter, Wales. Natural size. Olass 2. GEPHYREA. Marine Annelida without parajpodia and typically devoid of any trace of segnientation in the adult condition. The Cambrian genera referred to this class by Walcott differ in certain respects from the Kecent members, but with our available information the Position here assigned them seems most advisable. Two f amilies, (1) Ottoidae, with the genera Ottoia and Banffia, and (2) Pikaidae, including Pikaia (Fig. 217) and Oesia, all from the Middle Cambrian of British Columbia, are recognised. A quantity of supposed worm-borings, trails, impres- sions and otherobscureremains have been described from the older Paleozoic formations. The burrows have the form of straight or tortuous tunnels, and are sometimes hollow, but more commonly have been filled up by solid matrix. Various names have been applied to them, such as Scolithus, Arenicola, Histioderma, Planolifes, Diplocraterion, Spirocolex, Scolecoderma etc., but they are obviously incapable of precise determination. Arthrophycus Hall, originally described as a plant, Daedalus (including Vexillum Roualt) and Taonurus Fisher- ^^' \ t r. ^ and marginal pores (p), 5/i, (after B. and stde-plates (" pore-plates Ol Koemer). In Feiit- C). -B. Ambulacrum of Nuckocrinus. -j r\ i • ^ i-i n t Lettering as in ^ (after Roemer). vemites, Orophocnnus, and other genera, an addi- tional series of still smaller pieces, called the outer side-plates (^^ supplementär y pore-pMes" of Roemer), are placed between the side-plates and the walls of the radial sinus. Fentremites and Cryptoschisma have the entire upper surface of the lancet-plate exposed to view, and the side-plates are situated alongside of it in the a same plane. But in other forms the lancet-plate is wholly, or to a very large extent, concealed by the side-plates (Fig. 255, B), so that as a rule only a small space along the food-groove is visible. The sutures between the side-plates are indicated by shallow, horizontal furrows, which are con- tinued as superficial grooves over both halves of the lancet-plate as far as the median ambulacral groove. These crenulations, it should be noted, are frequently effaced in weathered specimens (Fig. 255, ^ and B). Small, pit-like depressions, or small tubercles, which are observable on the side-plates, indicate the places where the append- aeres or pinnules were f ormerlv attached. These are perfoctiy preserved pinnuies (afte/Meek 1 X- ^^ £ j ■ \. j. ^ ^ i and Worthen). only exceptionaily lound nitact, but when pre- served they completely conceal the ambulacral fields, and extend upward above the summit of the calyx (Fig. 256). They difFer considerably in length, even among species belonging to the same genus. They are jointed structures like miniature arms, uniserial as far as observed, but with ossicles sometimes wedge-shaped and interlocking to some extent from opposite sides, thus simulating a biserial arrangement. Whether they performed the function of discharging the ova, like the pinnules of Crinoids, can only be conjectured. Fig. 256. a, Pinnule of Fentremites, enlarged ; Orhitremites norwoodi (O. and S.) with l.,, BLASTOIDEA 165 The crenulations, or file-like markings across the ambulacra in Pentremites above iioted, are not mere surface ornamentation ; but the ridges constitute the sides, and the depressions the floor, of a series of small ducts leading from the pinnules, and forming lateral branches of the main ambulacral groove into which they discharge. These lateral ducts, as well as the main median groove of the ambulacrum, are, in well-preserved specimens, roofed over throughout the entire field by very minute alternating covering plates extend- ing all the way to the pinnules, and probably continuing along their ventral side. In this respect the structure of the ambulacral area has not been generally understood, and not heretofore correctly described. The arrange- ment of the side ducts, their discharge by a distinct curvature into the main groove, and their connection with the pinnules, leave no doubt that they were the closed food-grooves serving to conduct nutriment from the pinnules on toward the mouth. In most Blastoids the side-plates, or the outer side-plates when such are present, are pierced by marginal pores (or hydrospire pores), which communicate with the hydrospires. The pores are situated at the extreme outer margins of the plates, at the end of the lateral ridges, and alternate in position with the sockets of the pinnules. They are present in all forms having the hydrospires concealed within the calyx ; but are absent in the Codasteridae, in which the hydrospires are wholly or in part exposed on the outer surface. The hydrospires (Figs. 257, 258) are bundles of flattened, lamellar tubes, extending underneath the lancet- and side-plates, in a direction parallel with the boundaries of the ambulacral fields. They begin at the lower end of the ambulacra, and terminate in the hydrospire canals, of which the spiracles form the adoral apertures. When the spiracles are confiuent, the canals of adjacent groups of hydrospires enter the J; same opening. The hydrospires are suspended in the Vs^^the hdght «f^tue^^^^bi^^^^^^^ majority of forms along the walls of the body-cavity i, LaAcet-piate ; p, Pore-piates; (Pentremites, Fig. 257), being attached either to the ^' outer margins of the under lancet-plate or to a separate piece known as the hydrospire plate (Orbitremites, MesoUastus and Cryptohlastus). Pentremites has --^^..f Pentremites sulcatusSaj. Lower Carboniferous ; Illinois. Trans- erse section of calyx at about mps)^ Fig. 258. Transverse sections through the ambulacral fields, showing various forms of hydrospires. A, Orhitremites derhijemis. B, Orbitremites nonoooiU. C, Metablastus lineatus. D, OropTiocrinus verus. All sections enlarged (after Etheridge and Carpenter). from f our to nine hydrospires in each group ; Orhitremites two, or exceptionally one ; Troostocrinus and MesoUastus generally three, and Orophocrinus from five to seven (Fig. 258, A-D). In Phaenoschisma and Codaster (Fig. 260) the tubes 166 ECHINODERMATA— PELMATOZOA phylüm iv open externally by slits piercing the radials and deltoids and running parallel with the ambulacra. The functions of the hydrospires can only be surmised, but they are sup- posed to have served for respiration : they correspond doubtless to the pec- tinated rhombs and calycine pores of the Cystideans and to the respiratory pores of some Crinoids. It is probable that water was admitted to the hydrospire sacs through the marginal pores, and was discharged through the spiracles. Roemer and Forbes have suggested that the hydrospires may also have performed reproductive functions. Ludwig has called attention to the resemblance between the genital bursae of Ophiuroids and the slit-like spiracles in Orophocrinus ; his theory is that the hydrospires served both for purposes of respiration and for the discharge of genital products, a view which was also shared by Carpenter. The stem in Blastoids is preserved only in exceedingly rare instances. It is round, provided with a small axial canal, and composed of short joints, which apparently multiplied in a similar manner to that in the Crinoids. In Orophocrinus and Pentremites it has been traced for a length of 15 cm. without reaching the end; and in the latter form it has occasionally been found with a few, comparatively heavy cirri. A f ew genera, like Eleutherocrinus, are stemless. It has frequently been claimed, owing to the superficial resemblance of their ambulacral areas, that the Blastoids and Echinoids are mutually related ; but such presumptions are founded upon a total misconception of the value of external characters. The construction of the calyx, the presence of pin- nules, and the stemmed condition, are features which identify them unmis- takably as Pelmatozoa ; and their nearest relatives under this group are the Cystideans. The parallelism between the ambulacral fields of the one class and the recumbent arms, apparently soldered on to the calyx of the other, is self-evident. The hydrospires of Blastoids correspond to the pore-rhombs of Cystideans, as has already been remarked ; and the position of the mouth and anus is the same in both types. The Blastoids constitute a peculiar, but, on the whole, a very well-defined group, which is now regarded as of equal rank with the Crinoids and Cystids. : The earlier forms occurring in the Ordovician are primitive, representing transitions from ancestors of Cystid type, and having the characters of the two groups intermingled in varying degrees. In one genus, Asterohlastus, the presence of diplopores and lack of hydrospires are correlated with the presence of the Blastoid ambulacrum together with its bordering pinnules, and more strongly developed basals and radials. In another, Blastoidocrinus, the diplo- pores are replaced by hydrospires, thus further strengthening a line of development which becomes thoroughly established in the Silurian with the genus Troostocrinus. Several genera are represented in the Devonian, both of Europe and America. But the climacteric of Blastoid development takes place in the Lower Carboniferous of North America ; some of the beds of the Kaskaskia Group are densely charged with their remains, which, as a rule, are excellently preserved. They occur sparsely in the Upper Carboniferous and Permian of western America and the island of Timor, but above this horizon no traces of Blastoids have as yet been discovered. Nineteen genera, comprising upward of 120 species were recognised by Etheridge and Carpenter in their mono- graph of 1886, and a few have been added since. OLASS II BLASTOIDEA 167 The last-named authors subdivided the Blastoids into Begulares and Irreguläres, an arrangement representing incidental Variation rather than any broad morphological differentiation. That presented by Bather in Part III. of Lankester's Zoology, 1899, appears to be more logical, and is followed in principle here. By separating the typical Blastoids, in which the character- istic calyx plates have become fixed at a small and definite number, from the earlier forms which have not attained that structure, two main divisions may be recognised, viz. : ProtoUastoidea and EuUastoidea. Hudson, whose admirable studies upon Blastoidocrinus have thoroughly elucidated that hitherto obscure type, has suggested a third, FaraUastoidea, to express the differences shown by his researches between it and the other Ordovician forms. As the general division is a somewhat arbitrary one at best, it is thought that these differences are sufficiently emphasised by the family diagnosis. Order 1. PROTOBLASTOIDBA Bather (einend.) Calyx plates numerous, not limited to a definite numher. Family 1. Asteroblastidae. Blastoidea with calyx plates indefinitely arranged ahove basals and radials, and having, along wiih pentamerous ambulacra ami marginal hrachioles, diplopores and pore-plate, hut no hydrospires. Ordovician. Asterohlastus Eichw. (Fig. 259). Calyx gemmiform, pentagonal, pedunculate, and- composed o£ numerous rigidly united plates which are perforated by conjugate pores. Upper surface marked by five fig. 259. lircTP r»ptnlm'rl nr efpllnfp araaa wVnVli Asterohlastus stellatus Eichw&lA. Ordovician; large petalOia Or Stellate areas Wtlicn Puikowa, Russia. Natural size (after Schmidt). are occupied by alternating plates, and bordered by sockets for the attachment of brachioles. Ordovician ; Russia. Family 2. Blastoidocrinidae Bather (emend.). Calyx plates more definitely arranged in four ciixlets, without diplopores or pore- plate, hut with hydrospires present. Ordovician. Blastoidocrinus Billings (emend. Hudson). Calyx pentagonal, composed of four circlets of principal plates, viz. : (1) basals (number unknown) ; (2) radials with angular distal face, followed by (3) two large plates called bi- brachials, with numerous interbrachials in each interradius ; and (4) very large triangulär deltoids, with hydrospire-slits at their lower margins Adambulacrals, heavy covering plates, and some additional plates in the oral portion. Ambulacra large, bordered with numerous brachioles or pinnules. Base invaginate, with strong column occupying the concavity. Ordovician (Chazy Group) ; Canada and New York. Order 2. EUBLASTOIDEA Bather. Calyx plates limited to a definite numher of ahout thirteen. Hydrospires always present. 168 ECHINODEEMATA— PELMATOZOA PHYLUM IV Family 1. Oodasteridae Etheridge and Carpenter. Base usually well developed, and sometimes very long. Amhulacra without marginal pores. Hydrospire-folds Coming to the surface of the radial sinus. Hydrospire-slits either wholly exposed, piercing the calyx plates along the sides of the radial sinuses, or restrided portions of them remain open as spiracles, white the remaining parts are concealed hy the amhulacra. Devonian and Lower Carboniferous. Codaster M'Coy (Codonaster Koemer; Heteroschisma WsLchsm.) (Fig. 260). Calyx inverted, conical or pyramidal. Upper face broad, truncate or gently convex ; section, as a rule, distinctly pentagonal. Basais forming a conical or triangulär cup, usually deep. Kadials large, their limbs bent inward horizon- tally, to assist in forming the truncated upper face of the calyx, and never deeply excavated by the sinuses. Deltoids wholly confined to the upper face, as are also the amhulacra. The latter are petaloid, or narrow and linear ; lancet-plate, as a rule, deeply excavated for the side-plates. Spiracles absent, Fig. 2Ü0. Codaster acutus M'Coy. Lower Carbonifer- ous ; Derbyshire. A, Side-view of calyx. B, Base. C, Ventral aspect, enlarged (after Roemer). Fig. 261. Oropliocrinus stelliformis (O. and S.). Lower Carboniferous ; Burlington, Iowa. A, Calyx and base of the natural size. B, Ventral surface en- larged (after Meek and Worthen). hydrospires pendent, arranged in eight groups, two in each of the four regulär interrays, but wanting in the anal one. The tubes open externally by a variable number of elongated slits, which are separated by intervening ridges ; one or more of them may be partially concealed by the overlapping side-plates. Anus large, ovate or rhombic, and piercing the posterior deltoid. Ornament consisting of fine lines arranged parallel to the margins of the plates. Silurian to Lower Carboniferous; Europe and North America. Fhaenoschisma E. and C. (Fig. 254, B). Calyx resembling that of Codaster in general form, but with ten groups of hydrospires instead of eight. Radiais bear each three more or less distinct folds diverging from the lip; sinuses wide and deep, generally with steep sides. Deltoids small, confined to the truncated upper face of the calyx. Lancet-plates in all but one species (P. caryophyllatum) concealed by the side-plates ; outer side-plates very small. Spiracles rarely present. Hydrospires pendent, and opening externally by a series of elongate slits with intervening ridges, distributed in sub-parallel Order on the sloping sides of all the radial sinuses. The slits are only partially covered by the ambulacral plates, and are sometimes visible for their entire length. Lower Devonian ; Spain. Lower Carboniferous ; Europe and North America. CLASS II BLASTOIt)EA 169 Cryptoschisma E, and C. Calyx elongated, with a broad, fiat, truncated Upper face. Radial siiiuses wide and open, their sloping sides pierced by hydrospire slits, which are completely concealed by broad, petaloid ambulacra. Spiracles small, single or more rarely double ; in the latter case the posterior pair are confluent with the anus. Represented by the solitary species C. schultzi d'Archiac and de Vern. Lower Devonian ; Spain, Orophocrinus v. Seebach {Dimorphocrinus d'Orb. ; Codonites M. and W.), (Figs. 258 I), 261). Calyx balloon-shaped to truncate ob-pyramidal, with more or less concave upper face. Section distinctly pentagonal or stellate. Ambulacra narrow, linear to sub-petaloid. Deltoids generally visible in side- view, the posterior one wider than the others. Spiracles ten, varying from wide clefts along the sides of the ambulacra to narrow slits at their upper ends ; the posterior pair separate from the anus. Hydrospire-slits almost completely concealed, being concentrated at the bottom of the radial sinuses. Stem round, composed of short, nearly equal joints. Pinnules extending to nearly twice the height of the calyx, of uniform thickness throughout, and composed gf sharply cuneate pieces interlocking from opposite sides ; ventral furrow wide, and covered by small pieces. Lower Carboniferous ; Britain, Belgium and North America (Kinderhook and Burlington Groups). Family 2. Pentremitidae d'Orbigny. Base usually convex, and often much elongated. Spiracles five, hut sometimes more or less completely divided by a median septum, and bounded proximally by the upper- most side-plates. Lancet-plate either entirely visible or partially covered by side-plates which extend to the margins of the ambulacra. Hydrospires concentrated at the lowest part of the radial sinus. Devonian and Lower Carbon- iferous. Fentremites Say (Figs. 254-7, 262-3). Calyx usually ovate or pyriform, with elon- gate, sub-truncate base. Ambulacra broad, sub-petaloid. Lancet-plate wholly exposed, and resting below on an under lancet-plate. Side-plates and outer side-plates numerous, the former abutting against the edges of the lancet-plates. Hydrospires three to nine; spiracles single, or occasionally double ; the two of the posterior side confluent with the anus, and forming with it a single large orifice. Oral centre surmounted by numerous spines, placed closely against one another so as to form a pyramid, which completely Covers the summit and the greater portion of the Lower Carbon • 1 1 1 • 1 T iferous ; III spiracles. Lxcessively abundant m the Lower (Nat size). Carboniferous of North America (Burlington to St. Louis and Kaskaskia Groups), but not identified in Europe Defrance, and P. pyriformis Say, are the most familiär species. Pentremitidea d'Orb. Calyx clavate-pyramidal, with elongate, usually conical base, and truncate or convex upper face. Aml)ulacra narrow, short ; ■ Fig. 2C2. Pentremites (jodoni Defr. Fio, 263. Pentremites svlcatus Lower Carbon- iferous; 111.; A, Summit aspect. B, base. Roemer. P. godoni 170 ECHINODERMATA— PELMATOZOA PHYLUM IV lancet-plate more or less completely concealed by side-plates. Deltoids very small, generally confined to the upper face of the calyx, and seldom visible in a side-view. Spiracles and hydrospires as in the preceding. Lower and Middle Devonian ; Eifel, Ardennes, Spain, Great Britain and North America (Hamilton Group of Indiana, Michigan, Canada). P. pailletti de Vern.; F. eifelianus Roemer ; F. davatus Schnitze. Family 3. Troostoblastidae Etheridge and Carpenter. Calyx elongate. Ambulacra narrow, linear, deeply impressed, descending out- ward from the summit. Deltoids confined to the narrow upper end, rarely visible A externally, except the posterior one in Troostocrinus. Lancet- plate entirely concealed by side-plates. Spiracles distinct, repre- sented by lineal slits at the sides of the deltoid ridge, and bounded by deltoids and lancet-plates, but not by side-plates. Silurian to Lower Carboniferous. Troostocrinus Shum. (Clavaeblastus Hambach) (Fig. 264). Calyx narrow, elongate, somewhat fusiform, with contracted, subtruncate, or slightly convex upper face. Ambulacra short. The four anterior deltoids overlapped by the radial limbs ; the posterior one much larger than the rest, and appearing externally. Posterior spiracles conflueht with the anus. Silurian (Niagara Group) ; North America. Metablastus E. and C. (Fig. 258, C). Like the preced- ing, but all the deltoids equal, and the two posterior spiracles not confluent with the anus. Spiracle slits ten in number ; hydrospires four to each side of an am- bulacrum. Devonian to Lower Carboniferous (Keokuk Group) ; Europe and North America. Tricoelocrinus M. and W. (Saccoblastus Hambach). Calyx ir an+sp Fig. 264. Troostocrinus reinwarm pyramidal, broadest below and narrowing upwards ; when (Troost). Silurian nessee.^, Calyx from anal ggen from abovc or below stroudv pentaeronal in outline, side. B, Snmmit aspect. . .. i.ii r ^ (Z, Deltoid ; ir, Deltoid of owing to the projectmg and carinated character oi the anal side (after E. and C). j-i t\ ii. -j n i,i radials. Deltoids small ; ambulacra long, and extremely narrow. Spiracles ten, distinct ; anus large. Hydrospires small, enclosed within the substance of the forked plates. Lower Carboniferous (Warsaw Group) ; North America. Family 4. Eleutherocrinidae Bather. Elongate, stemless, asymmetrical, with four narrow ambulacra ; fifth ambulacrum shortened and widened. Hydrospires not con- centrated. Devonian. Eleutherocrinus Shumard and Yandell (Fig. 265). Devonian (Hamilton Group) ; Indiana, Kentucky, New York and Canada Fk;. 'Mb. Eleutherocrinus casse- dnyi (Shum. and Yand.). Lower Devonian ; Ky. Ventral surface, -/i (after E. and C). Family 5. Nucleocrinidae Bather. Calyx usually globular or ovoidal, with flattened or concave base, and linear ambulacra extending the whole length of the calyx. Spiracles distinctly double, and CLASS II BLASTOIDEA 171 chiefly formed hy the apposition of notches in the lancet-plate and deltoids. Devonian and Carboniferous. Nucleocrinus Conrad {Elaeacrinus Koeiner ; Olivanites Troost) (Fig. 266). Basais small, inconspicuous, sometimes hidden within the columnar cavity. Radiais small, with very short limbs. Deltoids greatly enlarged and elongated, forming over two-thirds of the entire calyx ; the posterior one wider than the others, and divided by a large anal-plate. Lancet-plate exceedingly long and narrow, partly exposed. Side-plates numerous ; hydrospires two on each side of the ambulacra. Summit covered by comparatively large orals, asymmetrically arranged and forming a flattened disk which completely closes the peristome. Devonian (Onondaga and Hamilton Groups) ; Indiana, Michigan, New York. Schizoblastus E. and C. (CribroUastus Ham- bach). Calyx resembling that of Orhitremites in form. Basais almost always confined to the lower face of the calyx ; deltoids of variable Size, but always Visible in a side-view. Hydro- Devonian ; Columbus, O.(afterRoemer). . ' / ^ y ^ A, Side-view of calyx. B, Base. C, Spn*es one to four to each ambulacrum. bpiracles ventral surface. D, Same enlarged. small, slit-like, placed between the lancet-plates and deltoid ridges ; the posterior pair sometimes confluent with the anus. Lower Carboniferous ; Ireland and North America (Kinderhook to Keokuk Groups); Permian, Timor. Fio. 266. Nucleocrinus verneuili (Troost). Lower Family 6. Orbitremitidae Bather. Calyx globular or owidal, with flattened or concave base, and long linear ambulacra. Spiracles five, piercing the deltoids, or ten, grooving their lateral edges. Lower Carboniferous. Orhitremites Austin (Granatocrinus Hall ex Troost MS. ; Cidaroblastus and Globoblastus Hambach) (Figs. 263 J, 258 A). Calyx ovate to globose. Lower face from slightly concave to deeply funnel-shaped ; interradial areas more or less depressed. Basais small, generally concealed in the central columnar cavity. Radiais very variable in size, often long, and invariably turned in below to assist in forming the base. Deltoids also variable ; usually unequally rhombic, but sometimes triangulär ; the anal deltoid f requently difFering from the others. Ambulacra nearly parallel-sided, always impressed within the sinuses at their proximal ends. Lancet-plates narrow, not Alling the sinuses, and more or less exposed throughout two-thirds of the ambulacra. Side-plates transversely elongated ; outer side-plates generally well developed. Hydro- spires pendent, usually but two or three folds on each side of an ambulacrum ; the inner one forming a well-defined hydrospire-plate. Spiracles five, piercing the apices of the deltoids. Posterior spiracle larger, including the anus. Summit closed by minute pieces which rarely exhibit any definite arrange- ment. Lower Carboniferous; England and North America (Burlington Group), (?) Australia. 172 ECHINODERMATA— PELMATOZOA phylum iv Cryptohlastus E. and C. Calyx sub-globose, with a flatteried or slightly hollowed base. Basais and deltoids small. Lancet-plate separated from the radials by a hydrospire-plate, which does not extend above the radio-deltoid suture ; but above this line the lancet-plate meets the deltoids without leaving any hydrospire-pores. Spiracles round, distinctly double at four of the sides, but those of the posterior side confluent with each other and with the anus. Summit covered by numerous, irregularly arranged small pieces. Lower Carboniferous (Burlington Group) ; North America. HeteroUastus E. and C. Eesembling the preceding in form and proportion of its component parts. The proximal ends of the deltoids produced in short spine-like processes, at the base of which minute lateral openings, one to each deltoid, are visible. These openings lead into gutter-like Channels excavated in the substance of the plates for the reception of the proximal ends of the two hydroBpire-canals. Eadial sinuses wide, their edges sloping gently downwards to the slightly petaloid ambulacra. Lower Carboniferous ; England and (?) North America. Mesoblastus E. and C. Calyx ovoid to globose, with concave to protuberant base. Eadials long, deltoids small, short, unequally rhombic. Ambulacra very narrow, extending to the base. Spiracles, as a rule, distinctly double, but sometimes incompletely divided. Lancet-plate entirely, or for the most part, concealed by side-plates. Lower Carboniferous ; Belgium, England, (?) North America and Australia. Acentrotremites E. and C. Calyx elliptical, with broad pentagonal lower face. Radials large, taking up three-fourths of the height of the calyx. Deltoids unequally rhombic, each notched by two spiracles at the ends of the radio-deltoid suture. Anal opening situated close to the summit in the posterior deltoid. Ambulacral edges of the deltoids without hydrospire-pores. Lower Carboniferous ; England. ' CarpenteroUastus and Lophoblastus Rowley. Lower Carboniferous (Kinder- hook and Burlington Croups) ; North America. (?) NymphaeoUastus von Beetz. Lower Carboniferous ; Russia. Family 7. Pentephyllidae Bather. Calyx stemless and sub-pentagonal ; radials asymmetrical. Ambulacra linear, extending dotun to the base ; one shorter than the rest. Carboniferous. Pentephyllum Haughton. Carboniferous ; Ireland. Family 8. Zygocrinidae Bather. Stemless. Calyx depressed, asymmetrical, quadrilobate. Four ambulacra between the lobes, accompanied by a single hydrospire on either side ; fifth ambulacrum shortened and widened. Zygocrinus Bronn (Astrocrinus Austin, non Conrad nee Münster). Lower Carboniferous ; Great Britain. [The text for the group Blastoidea in the present work has been revised by Mr. Frank Springer, of Las Vegas, New Mexico, and Washington, D.C.— Epttor,] CLAss III CRINOIDEA 173 Olass 3. ORINOIDEA Miller. Sea-lilies. ' (Brachiata Bronn ; Adinoidea F. Roemer.) Usaally long-stalked, more rarely non-pedunculate and sessile, frequently free- swimming Pelmatozoa, with calyx composed of regularly arranged plates, and provided with well-developed movahle arms. The Crinoid organism consists of three principal elements — calyx, arms and stalk. The calyx and arms together are sometimes spoken of as the crown, as contrasted with the column (also called stem or stalk). 1. The Calyx. — The calyx has usually the form of a cup-shaped, bowl- shaped, or globular capsule, within which the more important organs are enclosed. Its lower (dorsal or abactinal) surface commonly rests upon a column (Fig. 267) ; but in some forms it is attached directly by the base, and in rare instances it is free. The superior (ventral or actinal) surface is either membranous or plated ; it carries the mouth and ambulacral grooves, and hence is homologous with the under side of a star-fish or sea-urchin. As a rule, only the inferior and lateral portions (dorsal cup) of the calyx are visible, owing to the concealment of the summit by the arms. The cup is constituted of two or more circlets of plates, which are uniformly oriented with reference to the ambulacral organs. a. By the hase is understood the one or two circlets of plates intervening between the topmost Joint of the column and the first cycle of radially situated 1 Literatnre : Miller, J.