Trilobites

1

J

•J

T h e University o f C h i c a g o Press, C h i c a g o 6 0 6 3 7 T h e University of C h i c a g o Press, Ltd., L o n d o n ©

1975,

1 9 9 3 by the University of C h i c a g o

All rights reserved. Published 1 9 9 3 Paperback edition

1995

Printed in the U n i t e d States of A m e r i c a 0 2 01

00 99 98 97

3 4 5

ISBN: 0 - 2 2 6 - 4 7 4 5 1 - 8

(cloth)

ISBN:

(ppbk)

0-226-47452-6

Library of Congress Levi-Setti,

Cataloging-in-Publication

Data

Riccardo

T r i l o b i t e s / R i c c a r d o L e v i - S e t t i . — 2 n d ed. p.

cm.

I n c l u d e s b i b l i o g r a p h i c a l references a n d i n d e x . 1. T r i l o b i t e s — P i c t o r i a l w o r k s . QE821.L46

I. T i t l e

1993

565'.393—dc20

92—38716 CIP

© T h e p a p e r used i n this p u b l i c a t i o n m e e t s t h e m i n i m u m r e q u i r e m e n t s o f t h e A m e r i c a n N a t i o n a l S t a n d a r d for I n f o r m a t i o n S c i e n c e s — P e r m a n e n c e o f P a p e r for Printed Library Materials, A N S I Z 3 9 . 4 8 — 1 9 8 4 .

CONTENTS

PREFACE

VII

ACKNOWLEDGMENTS 1

BUTTERFLIES

2

THE

ARTHROPODA

3

THE

TRILOBITA

IX

OF

THE

SEAS

l

3 7

3 . 1 Introduction to Trilobite M o r p h o l o g y 7 / 3 . 2 Appendages and Internal A n a t o m y 1 8 / 3 . 3 T h e Eyes o f Trilobites 2 9 / 3 . 4 E n r o l l m e n t 7 4 / 3 . 5 Life Habits 7 8 / 3 . 6 Trilobite Classification 82 / A p p e n d i x A: A Case History: T h e G i a n t Trilobites of N e w f o u n d l a n d 87 / A p p e n d i x B: Photographic

Techniques and Specimen Preparation

105

4

ATLAS

OF

TRILOBITE

PHOTOGRAPHS

109

4 . 1 C a m b r i a n Families 1 1 3 / 4 . 2 O r d o v i c i a n Families 2 0 1 / 4 . 3 Silurian Families 2 4 7 / 4 . 4 D e v o n i a n Families 2 8 0 / 4 . 5 Mississippian Families 3 2 3 / 4 . 6 Pennsylvanian Families 3 2 6 / 4 . 7 Permian Families 3 2 6

REFERENCES INDEX

TO

327 GENERA

337

In fond memory of MY MOTHER GlLOA, who first mentioned to me the existence of curious extinct animals called trilobites, and that of MY FATHER, P AO L 0, who supported and encouraged my interest in rocks and physics instruments since my early childhood.

PREFACE

Trilobites tell me of ancient marine shores teeming with budding life, when silence was only broken by the wind, the breaking of the waves, or by the thunder of storms and volcanoes. The struggle for survival already had its toll in the seas, but only natural laws and events determined the fate of evolving life forms. No footprints were to be found on those shores, as life had not yet conquered land. Genocide had not been invented as yet, and the threat to life on Earth resided only with the comets and asteroids. All fossils are, in a way, time capsules that can transport our imagination to unseen shores, lost in the sea of eons that preceded us. The time of trilobites is unimaginably far away, and yet, with relatively little effort, we can dig out these messengers of our past and hold them in our hand. And, if we learn the language, we can read their message. I like to dig for trilobites. It is time travel and, at the same time, an addictive treasure hunt. I started doing this as an escape from my addiction to physics. Little did I know that I would encounter wonderful physics in the eyes of the trilobites—and other excitements of discovery. I also like photography, and trilobites provided an endless source of form and composition. Thus the origins of this book, primarily a personal account of my involvement with trilobites. The material presented in the first edition of this atlas, published in 1975, was accumulated over fifteen years of digging, preparing, camera work, and printing. Already at that time, my wish had been to share some of my findings with fellow scientists and students, while realizing that my pictures could bring excitement to a broader audience, of fossil amateur collectors and of naturalists at large. My desire to reach both groups of readers resulted in a compromise that would include a minimum of necessary technical coverage, letting my pictures do most of the talking.

VII

Much as it has been my experience in teaching introductory physics to a large class of either science or nonscience majors, it is extremely difficult to strike the middle ground that will enthrall some while awakening the interest of others. My 1975 atlas elicited responses and book reviews that ranged over a broad spectrum, many of them gratifying, some castigating, or both. These reviews have been of great help to me in this present endeavor. Another fifteen years have elapsed, and some of the issues that were raised and incompletely resolved then have now been clarified. The urge to communicate once more with my previous audience, and to involve a new one, has become imperative. The second edition of the atlas perpetuates the unavoidable dilemma faced earlier, concerning the technical level of presentation. With the aid of previous criticism, many former oversights and mistakes have been taken care of and certain pitfalls avoided. Thus the controversial issue of trilobite classification has been defused by presenting the

VIII

Preface

atlas materia] as it appears in the geologic column, grouping the related trilobite forms into families and subfamilies only. Although the original format and scope of the introductory text has been retained, its content has been updated to reflect the latest findings and lines of thought concerning the position of the trilobites in the early animal kingdom. Even if with some partiality, I have refrained from showing too many examples of my own trilobite finds. Instead, I attempted to cast my net farther with the help of friends who permitted my illustration of historical type specimens from museum collections, as well as of the treasures from their private collections. It is my hope that those trilobite enthusiasts familiar with the first edition of my atlas will want to find out how my thoughts have evolved and what has been added and enhanced in this revised edition. I also hope that, through these images and thoughts, a new generation of readers may be captivated and awed by this lost world of trilobites and by its message.

ACKNOWLEDGMENTS

I wish to express my gratitude to a great number of persons who over the years have contributed encouragement, inspiration, and practical help. I am particularly indebted to the late Dr. Eugene Richardson and to Matthew Nitecki of the Field Museum of Natural History, Chicago, for making the museum collections available for the selection of specimens to be used in this work, for the loan of those specimens, and for many helpful discussions. A great debt of gratitude is due to Dr. Jan Bergstrom who has supported the development of this work through the years with encouragement and inspiration. While I was struggling with the first edition of this book, Dr. Bergstrom, then at the University of Lund, helped me with much-needed information regarding the classification and life habits of trilobites. We then collaborated in fieldwork in Newfoundland and in the writing of a manuscript about the Paradoxides fauna mentioned in Appendix A of my first

edition of this book. It has been a treasured experience to unscramble together the many puzzles presented by our findings. A field trip through classical Swedish trilobite localities, guided by Dr. Bergstrom, has been the highlight of a trip to Sweden some ten years ago. As this second edition was taking shape in recent years, Dr. Bergstrom, now at the Swedish Museum of Natural History in Stockholm, did help me once more with much dedication. My views concerning trilobites were updated and corrected through a critical reading of the manuscript and much appreciated advice. During a recent visit to Stockholm, Dr. Bergstrom gave me the opportunity to peruse the trilobite collections and to photograph several type specimens of historic value at the Swedish Museum and at the Geological Survey of Sweden and the Paleontological Institute of the University in Uppsala. The section on trilobites eyes would not have existed without the

contributions by Dr. E. N. K. Clarkson of the Grant Institute of Geology in Edinburgh. For his magnificent photographs and the communication of many of his results, a great deal of appreciation is due. It has been most rewarding for me to collaborate with Dr. Clarkson on the writing of a manuscript on the optics of the phacopid eye lenses. This edition contains the conclusion of this investigation, which was still in progress at the time the first edition of the atlas was published. My visit to the Grant Institute of Geology in Edinburgh, and memorable field trips to the Pentland Hills and Girvan guided by Dr. Clarkson, have been a much appreciated opportunity. Through the courtesy of Dr. J. Cisne, I had access to the collection of Beecher's Utica shale trilobites as described in section 3.2. In addition to the loan of many of the trilobites, many X-ray negatives were most generously contributed by Dr. Cisne, to whom I am deeply grateful. IX

>

I wish to thank Dr. Bernhard Kummel of Harvard's Museum of Comparative Zoology for allowing my perusal of the museum collection and for the contribution of some of the most beautiful trilobites in the entire atlas, from Bohemia in particular. I am also grateful to Dr. N. Eldredge of the American Museum of Natural History and to Dr. K. Towe of the Smithsonian Institution for their discussions and useful communications of their work. Dr. A. M. Ziegler at the University of Chicago has helped on many occasions with his stimulating criticism. Further appreciation is due the superintendents of many quarries for their permission to collect fossils; particular thanks go to Mr. H. Nester of the Consumer's Company Quarry at McCook, Illinois; Mr. J. Riordan of the Lehigh Stone Company, Lehigh, Illinois; and the Medusa Cement Company at Sylvania, Ohio. Many more persons have helpedwith permission to collect, to photograph specimens from their collections, with loans of specimens, and with precious gifts. In preparing this second edition, I was privileged to have access to the extraordinary trilobite collection of David C. Rilling, M.D., of Sellersville,

X

Acknowledgments

Pennsylvania. Many exceptional specimens were photographed on location with his kind assistance. Another mutual friend, Pio Pezzi, M.D., of Huntington Valley, Pennsylvania, has loaned several valued specimens. A word of appreciation is also due a number of fossil dealers and trilobite preparators, from whom many specimens have been obtained or loaned. Among these, Mr. Afton Fawcett of Hurricane, Utah, has provided many trilobites, collected and prepared with professional care. Robert Carroll of Ann Arbor, Michigan, did contribute with the superb preparation of several trilobites exhibiting extremely delicate structures. To be mentioned also are Dave Douglass of Evanston, Illinois; Michael Thomas of York, Pennsylvania; Leon Theisen of Hill City, South Dakota; Eddie Cole of Delta, Utah; Bruno Corti of Milan, Italy; and Franco Todde of Iglesias, Sardinia, Italy, for enabling me to portray specimens from their collections. I am particularly indebted to my sons Emile and Matteo, who since they were first able to climb boulders have very successfully competed with their father in finding countless trilobites. Once, many years ago, in Rome, I had the opportunity of meeting

Franco Rasetti. "Keep them sharp," he kept telling me, referring to the tools needed to free trilobites from their matrix. The memory of that meeting, and of much constructive advice, is still cherished. When presented with the first edition of my atlas, he admitted surprise at finding out that so much could be said about Post-Cambrian trilobites. Those who know of Rasetti's contributions to physics and paleontology will understand the reasons for a particular devotion to this great scientist on the part of another trilobiteloving physicist. Trilobites have pervaded much of my personal life, vacation time, as well as my home environment. The second edition of my book would not have materialized without the devoted help and encouragement of my wife, Nika, who willingly endured the hardships of many field trips and countless hours of digging. For her patience and endurance, in the field and at home, I am deeply indebted. It is a pleasure to acknowledge the help of several students, apprentices in the art of photographic printing. To be mentioned are Margarita Garcia, Marian Harris, Claudine Malik, and Tara Shochet.

BUTTERFLIES

OF

THE

SEAS

Among the remains of early life on earth, the fossil record we find buried in ancient sedimentary rocks bears evidence of an extraordinary group of marine creature, the trilobites. The position of these invertebrates in the evolution of the animal kingdom is extraordinary because of their early ascent to a high level of functional complexity, described in fascinating detail by their persistent and ubiquitous fossil remains. Trilobites could see their immediate environment with amazingly sophisticated optical devices in the form of large composite eyes, the first use of optics coupled with sensory perception in nature. As a unique feat in the history of life, their eye lenses were shaped to correct for optical aberrations, with design identical to that proposed (quite independently of any knowledge of trilobites) by Descartes and Huygens. Although we can only hold the petrified remains of this long-extinct form of life, what was preserved of trilobites constitutes a record of immediate and striking impact, a still life we can interpret and recognize. It is this extravagantly rich still life I want to illustrate in this book, preserving as much as possible of the excitement of a voyage back in time to the dawn of life. Trilobites share with many other invertebrates in the drama of the "Cambrian explosion": the rapid evolution of amazingly sophisticated and diverse life forms at the beginning of the Cambrian period, some 570 million years ago. In a sort of chain reaction triggered by plant life through a large-scale photosynthetic release of oxygen in the atmosphere, the pattern and rate of biological evolution was tremendously accelerated. Energy-releasing, oxygen-burning processes became available to the evolution of animal life. The newly discovered powerhouse fostered experimentation with ever more demanding levels of body activity and multiplication of forms. During a relatively short (geologic) time, perhaps a few million years, different new schemes of bodybuilding emerged that involved the 1

construction of internal and external skeletons or exoskeletons. These could support increasingly complex body structures and functions. The genetic radiation of new life forms found coundess environments devoid of enemies in which to adapt and prosper. Amidst this revolution, trilobites evolved to appear suddenly in the fossil record, when they had already reached a full degree of development and differentiation. Compound eyes are already present in the earliest trilobites. The preservation of trilobites as fossils in their burial rock is related to the fact that the living animal possessed an already mineralized exoskeleton. Somehow, trilobites learned early in the game to metabolize calcium carbonate in building sturdy shields. This was most likely another consequence of the explosive increase in the level of oxygen in the atmosphere that made carbonate more available. Indeed, the fact that trilobites heralded the advent of advanced life forms in the fossil record can be regarded as a most peculiar event. Trilobites were not the only inhabitants of the primordial seas. Snapshots of the life forms that coexisted with early trilobites about 540 million years ago in the Middle Cambrian are provided by the soft-bodied fauna unusually preserved in the Burgess Shale of British Columbia. This assemblage reveals an extraordinary diversity of extravagant marine animals that left no record in ordinary sedimentary rocks due to the lack of hard body parts. An even older and similar record, dating at about 570 million years of age, the Lower Cambrian, has been recently discovered in Chengjiang (Yunnan Province), Southern China. Most of these peculiar life forms, even if never seen again, could be recognized as belonging to animal groups or phyla that carried through to recent times. The trilobites appear

2

Trilobites

as a minority of species in the Burgess and Chengjiang faunas. And yet, because of their mineralized exoskeleton, they have left with us the most compelling evidence of their remarkable functional complexity. The high point of the trilobites' differentiation of forms was reached toward the end of the Cambrian period, some 500 million years ago. They became extinct at the end of the Permian period, the borderline between the Paleozoic and Mesozoic eras, around 230 million years ago. Dinosaurs were still largely in the making at that time. Long before the age of fishes, which began about 400 million years ago, the trilobites had already survived evolutionary decimation by about 170 million years. The span of time which saw the birth, development, and disappearance of trilobites and their age is indeed staggering. However, the fact that their general appearance is not so dissimilar from that of living arthropods—the horseshoe crab, for example—makes trilobites less forbidding than some other forms of extinct life. The widespread abundance of trilobites, particularly during the Cambrian period, is evident from layers of sedimentary rocks which are occasionally coated with trilobite remains. The presence of particular genera in layers of determinate geologic age makes trilobites very important index fossils. The presence of identical forms in rocks of identical age and composition on locations now separated by oceans is telltale evidence of the drift of the continents on the earth's crust. The fascination of the trilobites' age, their role in early life on our planet, the optimization of their visual organs, the ingenuity of their life adaptation to the environment, and their value as geological markers explain why trilobites are important to both the professional scientist and the amateur fossil collector.

THE

ARTHROPODA

In the subdivision of the animal kingdom into major groups or phyla, the trilobites are included in the phylum Arthropoda. The history of these highly developed invertebrates begins abruptly with the beginning of the Cambrian period and extends to modern times. No animals with skeletons or shells are found among the well-preserved traces of Precambrian life forms, and since the hard parts of arthropods (and many other phyla) are an integral part of their organization, we must infer that they were not in existence as yet. Trilobites are amongst the oldest fossil arthropods. Familiar arthropods are the insects, scorpions, spiders, centipedes and millipedes, shrimps, lobsters, crabs, and others with jointed legs. Although they first inhabited the seas, the arthropods have adapted themselves most efficiently to every life habitat. The position of the Arthropoda, and of the Trilobita in particular, in the animal kingdom is illustrated in figure 1. Here the principal groupings into classes and subclasses are traced through, by vertical lines, the Phanerozoic time scale (the ensemble of geologic eras with a fossil record), from their first appearance to present times. Boxes indicate the higher ranking of classes, often related, into phyla. This representation is very schematic and does not show the variations in the number of species encompassed by each phylum as a function of time. In the histogram at the base of figure 1, however, we can appreciate a cumulative comparison of the relative abundance of known extinct and living species. The Arthropoda actually represent a much more prolific group of animals than is shown in such a histogram. In fact, the one million or more species of insects, representing seventy percent of all species of extinct and living animals, had to be removed from the plot so that the remnant fractions could become visible. This may come as a surprise to many, but from the point of view of a statistician of living species, we live today in the age of insects! It would be misleading to 3

FIGURE

1

C o m p o s i t i o n of the animal k i n g d o m

subclasses are connected by an e m p t y

figure indicates the percentages of living

through geologic time. O n l y the

box. T h e time scale indicated was

species, excluding insects (partially from

ptincipal phyla are indicated (names in a

adopted by the Geological Society of

Easton

box), along w i t h s o m e of the classes they

L o n d o n i n 1 9 6 4 (from K u m m e l 1 9 7 0 ) .

1 9 5 3 ; and from a survey by the author).

encompass (vertical lines). Related

T h e histogram at the b o t t o m of the

4

Trilobites

1 9 6 0 ; S h r o c k and Twenhofel

say that the early Paleozoic was the age of trilobites, even though their remains dominate the fossil record of that period. We now know that a surprising richness of other marine invertebrates constituted only of soft parts did coexist with the trilobites and that trilobites represented only a small fraction of a major proliferation of arthopods. The multitude of life forms that existed in the Middle Cambrian period is beautifully illustrated in the book The Burgess Shale by H. Whittington (1985) and more recently traced to the beginning of the Cambrian from the study of the Chengjiang fauna. A view of the latter by a group of collaborating Swedish and Chinese paleontologists (Chen et al. 1991; Xianguang et al. 1991) restricts the development of the principal arthropod groups and of most of the other advanced multicellular life forms to the very transition between the Precambrian and the Cambrian. Undoubtedly, the Cambrian explosion may well have been the most revolutionary and far reaching single event in the history of life. More accessible than their soft-bodied relatives, even if less abundant during their lifetime, the ubiquitous trilobites that we find as fossils provide us all with a telltale reminder of this special time. Evidently the organizational scheme of the arthopod body, perhaps because of its versatility, has proven most apt to secure survival of some of its offshoots. In its most primitive form, the body plan is modular or metameric, made up of a sequence of basically identical segments (serial homology) performing similar functions. The latter encompassed locomotion, food gathering, respiration, excretion, and reproduction. The body, made up of a sequence of segments, is generally of elongated form and exhibits bilateral symmetry about an axis carrying the digestive tract. One of the functions of the segments is that of articulating the body. Each segment also usually carries pairs of appendages which are, in turn, divided into jointed segments or podomeres (hence the name Arthropoda, from the Greek arthron = joint, podos = foot). The scheme is highly adaptive. Each block of segments can evolve to respond to specific needs or functions, without limiting the adaptation of other portion of the body

to perform different tasks. Such functional groups of segments are called tagmata (from the Greek tagma for a "body of soldiers"). Segmental specialization into tagmata or tagmosis leads to the fairly universal organization of the body into head, thorax, and abdomen. Thus the head tends to monopolize the sensory functions, including eyes, antennae, and related food-gathering and masticatory functions, while successive tagmata may encompass locomotory appendages and/or swimming appendages, etc. The dorsal portion of contiguous plates (tergites) may become fused into a shield (tergum), as that covering the head of most extant and extinct arthropods. In extreme cases, as in the horseshoe crab (Limulus), fusion occurs between head and thorax to form a solid shield called prosoma. Differentiation and specialization can also take place along individual podomeres, giving rise, for example, to antennulae, claws, etc.—the tools that are needed by the tagmata. The nervous system of the arthropods is highly developed, with pairs of ganglia in most segments, and a pair of nerve cords leading to a more or less developed brain. The interaction of the animal with the environment takes place by means of sensory organs, such as tactile antennulae and single or compound eyes. The latter, as we shall see in detail for the trilobites, can attain surprising complexity and optimization of function. Arthropods possess a circulatory system consisting of heart, arteries, and blood-return ducts. Respiration takes place through gills in the aquatic forms or through a network of tracheae in the air-breathing forms. The sexes are distinct, and eggs hatch either externally or internally. Growth takes many forms. We are familiar with the metamorphosis of the caterpillar into a moth or butterfly. The soft arthropod body is occasionally encased in a protective shield, or exoskeleton, made mostly of hardened proteins and chitin, which may be further hardened by calcium carbonate or phosphate. The presence of a hard, inextensible exoskeleton poses special problems to the growth of many arthropods. In such cases the increase in size takes place by ecdysis (molting), in which the hard cover is shed periodically, and a new, larger shield is constructed after each

The Arthropoda

J

{'A

step. Soft-shell crabs are examples of the stage of growth following ecdysis. For a period of time, the animal is protected only by a flexible, proteinaceous exoskeleton that will gradually harden by mineralization. Based on morphological and anatomical similarities, body segmentation in particular, it has been assumed for a long time that the arthropods may have been related to a common annelid ancestor way back in the Precambrian. During the last twenty years, a new understanding of animal phylogeny has developed, based on the mapping of sequences of nucleotide bases and aminoacids for different animal groups. Not only have the arthropods emerged as

6

Trilobites

biochemically different from the annelids, but the reliability of morphological considerations in establishing phylogenetic relationships has been proven misleading in several other cases. Most of the phyla that emerged as a result of the Cambrian explosion are thought to have originated as offshoots of a hypothetical phylum of sluglike animals devoid of true segmentation, called Procoelomata (Bergstrom 1989). Segmentation like that found in arthropods would have been attained from a state of false segmentation, commonly occurring in primitive animals. Several of the offshoots of the Procoelomata are believed to have developed similar morphological characters by convergent evolution.

THE

3.1

TRILOBITA

Introduction to Trilobite Morphology

The general description of the distinctive features of the Arthropoda given in chapter 2 applies to the Trilobita as well. Here, however, the picture is less complete than for the living arthropods, since, of course, our evidence must be based only on the fossil record. Several details of the anatomy and physiology of trilobites are inferred by analogy with what is known about the character of arthropods in general. As much as this analogy may be useful, however, it is important to notice that the evolution of trilobite morphology and life habits differed significantly from that of most arthropods. Throughout their long survival, trilobites preserved a relatively primitive body structure, characterized by serial homology manifested in the repetitive appearance of nearly identical segments and by a low degree of tagmosis. This generalized, primitive arthropod morphology is encountered in only a few modern descendants. By contrast, instead, most arthropod groups progressed toward special adaptation of groups of segments to perform functions such as locomotion, food gathering, defense and aggression, respiration and others. As we shall come to understand, the limited specialization of arthropod design in trilobites may be consistent with a predominantly nonpredatory life-style and with humble food-gathering habits. Trilobites were, on the average, small animals, two to seven centimeters long, with extremes ranging from three millimeters to about seventy centimeters. They lived in the sea, for their remains are always associated with those of marine animals (corals, brachiopods, cephalopods, and so forth). The commonly preserved portion of the body of trilobites is the dorsal shield, or carapace, made of sclerotized and mineralized protein. Most of the ventral part of the exoskeleton is generally not preserved, and we must infer that it consisted 7

of a cuticle made of nonmineralized protein. As a rule, all soft parts as well as appendages are also missing in the fossil trilobite. There are, however, exceptional forms of fossilization in which most of the soft pans have been preserved, and these few examples have given a powerful insight into the detailed anatomy of these extinct arthropods. We shall deal with such details in section 3.2, limiting this introduction to the more apparent characteristics of trilobite morphology and life habits. For the description of trilobite morphology and nomenclature we shall refer to the reconstruction of a trilobite in figure 2, representing the dorsal view of the carapace of Paradoxides gracilis (Boeck), a beautiful trilobite from the Middle Cambrian of Bohemia. Shown in plate 1 is the photograph of the original specimen, on which the reconstruction has been based. We recognize immediately the bilateral symmetry of the trilobite body, a characteristic that is shared with a large majority of animal groups, collectively described as Bilateria. Here the strongly convex axis takes the name of axial lobe and the two adjacent regions are called pleural lobes. The pleural lobes are separated from the axial lobe by two axial furrows. It is from this longitudinal trilobation (separation into three lobes) that the name "Trilobita" originated and not from the longitudinal subdivision of the body into the three regions of the cephalon, the thorax, and the pygidium, as is often erroneously supposed. The latter subdivision is in fact a characteristic that trilobites have in common with most arthropods. The reader interested in a more comprehensive coverage of the topics, which by necessity are mentioned only briefly in this context, is referred to theTreatise on Invertebrate Paleontology (Moore 1959).

3.1.1

The Cephalon

The head shield or cephalon resulted from the fusion of a number of tergites (five or seven) and often carries telltale memory of the original segmentation. It is the most significant and characteristic part of the trilobite morphology.

8

T h e Trilobita

The outline of the cephalon may be semicircular to ogival in its anterior portion, straight or gently curved at the posterior margin, where articulation with the first thoracic segment occurs. The lateral and anterior margin of the cephalon are inflected into the doublure, a narrow strip of the ventral side of the exoskeleton that is mineralized like the dorsal side. Attached to the anterior part of the doublure is a small shield called hypostoma, one of the few hard parts to be found on the underside of the trilobite. The angle between the backward-sloping lateral margin of the cephalon and the posterior margin is called the genal angle. This termination can be rounded or, as in our example in plate 1, prolonged into long genal spines. The axial lobe extends into the cephalon, where it takes the name of glabella. This can be a very convex bulging region, sometimes extending all the way to the anterior margin, or it can terminate earlier, defining a flat preglabellar field. The furrows may delineate an anterior lobe and several pairs of lateral glabellar lobes. This particular structure together with the occipital ring provides suggestive evidence of the original segmentation of the cephalic region and is apparent in the trilobite chosen for our reconstruction. The furrows which surround the glabella take the name of the particular area where they are located, as indicated in figure 2. (This is a general rule for furrows occurring elsewhere in the trilobite exoskeleton.) On the sides of the glabella we find the cheeks, which are split into

two regions, the free cheeks and the fixed cheeks, by the facial sutures. The assembly of the two fixed cheeks together with the glabella constitutes the cranidium. About the middle of the exterior margin of the fixed cheeks is a kidney-shaped elevation, the palpebral lobe. The visual surface of the eye is usually located on the outer slope of the palpebral lobe. Much will be said about the eyes of trilobites in section 3.3, so we shall not go into details here. There are several junctions, or sutures, between the various parts of the cephalon. The facial sutures mentioned above have been considered meaningful for classification until recently, when other overriding criteria have been developed. Based on the position of the facial sutures, the

FIGURE

2

Description o f trilobite terminology, (a) Dorsal view o f the c o m p l e t e exoskeleton. (b) V e n t r a l v i e w of cephalon. T h e trilobite represented is

Paradoxides gracilis

3.1

Introduction to

Trilobite Morphology

(Boeck).

9

PLATE

1

Dorsal view of a c o m p l e t e specimen

gracilis

of

Paradoxides

(Boeck), a M i d d l e

C a m b r i a n trilobite from the Jince Formation of Jince (Jinetz), Bohemia. T h e reconstruction in figure 2 schematizes the characters visible in this adult individual. Distortion due to tectonic shear is often present in the trilobites from this locality. In this photograph, the shear has been partially corrected by tilting the object relative to the camera viewing direction. Specimen w h i t e n e d with magnesium oxide, ( x l . 3 , RLS coll.).

10

The Trilobita

trilobites are termed opisthoparian if the suture cuts the posterior margin, proparian if the sutures run to the lateral margin. In the former case, the genal angle and spine are carried by the free cheeks, in the latter by the cranidium. Figure 2 shows an opisthoparian suture. The free cheeks often separate easily from the cranidium, and this feature greatly facilitated molting. A third group exists in which the suture terminates at the genal angle, termed gonatoparian (e.g., Calymene), and a fourth group without facial sutures (e.g., the Olenellida). As we shall see, the cephalon of trilobites has differentiated into a variety of forms characteristic of particular groups, and much of trilobite classification is based on cephalic features.

3.1.2

The Thorax

This is the section of the exoskeleton where separate tergites are articulated with each other to enable flexibility and enrollment. The number of thoracic segments varies between two and as many as forty-two, most frequently in the range of eight to fifteen. Each tergite consists of a center part—the axial ring—and two adjacent pleurae. These may terminate bluntly or may arch downward and backward into pleural spines of varied length. The pleurae are commonly sulcated by a pleural furrow, which may have served the function of strengthening the tergite. The articulation of one tergite with the other occurs through an anterior extension of the axial ring, which is inserted beneath the posterior margin of the next tergite. The interlocking mechanism between segments and the extent of rotation allowed evolved considerably between the earlier trilobite forms and more advanced ones. This function obviously determines the mode and capability of enrollment, which will be discussed in section 3.4.

3.1.3

The

Pygidium

In most trilobites, the posterior thoracic region or abdomen terminates in a shield called pygidium. This may result from the fusion (or lack of separation during growth) of a number

of tergites. Extreme variation in size occur for this portion of the carapace, and the extension of the pygidium seems related to the number of thoracic segments. In the example in figure 2, the pygidium is small-the trilobite is called micropygous-but the number of thoracic segments is relatively large (twenty for the example shown). Frequently, the pygidium reaches a size comparable to that of the cephalon {isopygous trilobites), as will be seen from many examples in the atlas. In such cases, the number of thoracic segments is usually small. Macropygous trilobites are provided with pygidium larger than the cephalon. The axial lobe clearly extends into the pygidial region in a great majority of trilobite species, so that the trilobation is usually preserved. It may extend to the posterior margin or terminate earlier. Evidence of the segmentation from which the pygidium was derived is to be found in the frequently observed ribbing of the axial lobe, resembling the articulated axial rings of the thorax. This causes the axial lobe to appear continuous from the thorax through the pygidial region. There are pleural regions resembling thoracic pleurae, sulcated by furrows. When the latter do not reach the lateral margin, a smooth border results. The marginal border is turned under to form a doublure, much as in the cephalic and thoracic regions.The shape and ornamentation of the pygidium may reach extravagant extremes. Marginal spines are often present, which may be related or not to the pygidial pleurae. Occasionally the pygidium may terminate with a long axial spine, as the continuation of the pygidial axis. Exceptionally, as in the Olenellidae, the pygidium is formed of a single tergite, shaped as a true telson.

3-1.4

Growth and Molting

For a number of trilobite species, the various stages of growth (ontogeny) from the larval to the adult form are known with great detail from the fossil record. Three major periods of growth are recognized. The protaspid period extends from the hatching of the egg to the first appearance on the singlepiece dorsal shield of a transverse suture, defining the cephalon and the so-called transitory pygidium. During this 3.1

Introduction

to

Trilobite

Morphology

11

period, a larval ridge may be the precursor of the axial lobe. The size of the protaspis is very small, typically 0.3 to 1.0 millimeters, and the lower limit for any species is clearly the upper limit for the size of the eggs of that species. During this period, the protaspis may develop features, such as marginal spines, which will disappear at a later stage. Observation of the fossil record of growth in the protaspis period has led to speculations concerning the affinities of trilobites with other arthropods, based on the notion that larval development may be a replay of ancestral history or phylogeny. The meraspid period is characterized by gradual separation of the cephalon from the transitory pygidium, by means of the progressive appearance of thoracic segments. Meraspid degrees correspond to the number of segments which made their appearance. The new segments originate at the thoracic-pygidial boundary. The number of molts required to complete the thorax does not necessarily correspond with the number of segments to be added to the thorax. The overall size of the trilobite may increase to more than ten times that of the protaspis. The meraspid period terminates when the thorax has reached the number of segments which characterize the adult individual. At this stage, the adult form, or the holaspidperiod, has been attained. Here the growth is continuous through many moltings, so that the size of the individual is correlated with its age. Several changes in the relative size of the various parts of the exoskeleton do occur throughout this period. The cephalon usually represents a larger fraction of the carapace in the early growth stages. The ontogeny of trilobites is beautifully illustrated in an article by H. B. Whittington in the Treatise of Invertebrate Paleontology (Moore 1959). The illustration of this aspect of trilobite life will be limited here to the presentation of a pictorial summary in plate 2. This picture contains evidence of some kind of trilobite nursery, containing examples of all stages of growth. The trilobite depicted is Elrathia kingii (Meek) from the famous Wheeler Formation of the Middle Cambrian of Utah. Although detail may be lost in the smallest specimens, this view gives a feeling for the range of sizes involved in trilobite growth. Present are several 12

The Trilobita

holaspid carapaces of various length, with their characteristic number of thirteen thoracic segments; at least one meraspid carapace with eight segments; and at least one protaspid shield about one millimeter long, showing the larval ridge quite distinctly. Molting is clearly an integral part of the growth process in trilobites. The most abundant fossil remains of trilobites are the disarticulated exuviae, which, on account of their relatively large surface and light weight, could be easily transported and concentrated by wave motions and currents. Plate 3 shows a slab of shale from the Collingswood Formation, Ontario, which is densely covered with a multilayered assembly of carapace fragments of Pseudogygites latimarginatus (Hall). Such accumulation is most likely due to selective concentration, as is often seen occurring with small bivalve shells on the seashore. In such occurrences, complete carapaces are very seldom found. Since trilobites molted many times during their growth, it follows that each individual left a multiple fossil record. In situations where the molting process occurred in a relatively undisturbed environment, exuviae may be found in the approximate posture in which they have been abandoned by the trilobite. There are at least two characteristic modes in which the exoskeleton was shed by the growing trilobite. In one of them, the phacopid mode, the cephalon separated from the thorax between the occipital ring and the first thoracic segment. The animal could crawl out of the old exoskeleton through such an opening, and in so doing would force the old cephalic shield to flip upside down. Plate 4 shows one of many examples, collected by the author, of exuviae shed by Phacops rana milleri Stewart, as found in the Devonian Silica Shale at Sylvania, Ohio. The cephalon is intact and usually is found in the immediate vicinity of the thorax-pygidium assembly. The latter occurs most often in the enrolled condition, showing that the integument must have contracted like a spring, after the former inhabitant crawled out. As mentioned previously concerning Arthropoda, the newly molted animal is provided with a soft exoskeleton, which would later harden through sclerotization and mineralization. Plate 5 shows a "soft-shelled

PLATE

2

V a r i o u s stages of g r o w t h o f the trilobite

Elrathia kingii

(Meek)

f r o m the W h e e l e r Formation, Middle Cambrian, Utah ( x 2 . 7 ) . ( R L S coll.; n o w a t F M N H . ) T h e larger trilobites represent the holaspid stage, a meraspid carapace w i t h eight segments is the center l o w e r m o s t trilobite, a protaspid shield is located just below the l o w e r m o s t complete trilobite on the right-hand side.

13

PLATE

3

Slab coated with disarticulated exuviae of the

trilobite

(Hall)

Pseudogygites

latimarginatus

from the C o l l i n g s w o o d

Formation, O r d o v i c i a n , C o l l i n g s w o o d , O n t a r i o ( x l . l ) . ( R L S coll.; n o w a t FMNH.)

1 4

T h e Trilobita

trilobite," also from the Silica shale. We are dealing here once again with Phacops rana milleri Stewart. This soft-shell condition is revealed by the fact that the carapace is considerably thinner than in the average specimen. This gives a translucent quality to the shield, particularly apparent in the pleural regions. The axial lobe seems to have thickened somewhat more and appears darker. By accident or not, this trilobite overlaps the exuviae of another trilobite (or its own?). The above description seems to ignore that we are dealing with a fossil and not with a living animal. However, here the fossilized exoskeleton is composed of the primary calcite that was originally present in the trilobite cuticle. Thus, the color differentiation that affects selected areas of the carapace in plate 5 may reflect incomplete mineralization of the exoskeleton as it existed at the time of burial.

Another pictorial view of phacopid molting is contained in plate 6. Here we deal with exuviae of Dalmanites verrucosus (Hall) from the Silurian Waldron Shale formation of Waldron, Indiana. In this example all trilobite parts are seen from the ventral side and are still partly imbedded within the shale matrix. The hypostoma appears displaced from its axial position, indicating opening of the hypostomal suture during molting. In the exuviae shown in plate 5, the hypostoma is missing altogether. Another frequent molting procedure is the olenid mode, in which the free cheeks separate from the cranidium at the facial sutures, enabling the molting trilobite to make its way out of the opening thus created. In this case, the facial sutures are functional, representing natural fracture lines.

PLATE

4

Exuviae of Phacops

rana milleri Stewart, from the Silica Shale, Devonian, Ohio

Sylvania,

(x4.5).

(RLS

coll.; n o w at F M N H . )

3.1

Introduction

to

Trilobite

Morphology

15

PLATE

5

An example of "soft-shelled trilobite" ( x 2 . 7 ) . D u e to the unusually thin and translucent carapace, this example of

Phacops rana milleri

Stewart is

interpreted as representing a newly molted animal. Silica Shale, D e v o n i a n , Sylvania, O h i o . (RLS coll.; n o w at FMNH.)

16

The Trilobita

PLATE

6

Exuviae of Dalmanites

verrucosus

(Hall),

seen

from the ventral side ( x 3 . 8 ) . Silurian, W a l d r o n Shale, W a l d r o n , Indiana. ( R L S coll.; n o w at F M N H . ) N o t e the displaced h y p o s t o m a .

3.1

Introduction

to

Trilobite

Morphology

17

In spite of the predominance of exuviae in the fossil record of trilobites, an animal would occasionally be buried intact, as the majority of the pictures in the atlas section will illustrate.

3.2

Appendages and Internal Anatomy

Although the more commonly preserved part of the trilobite body is the hard, mineralized carapace, there are a few fossil deposits that have yielded completely preserved bodies, from which, by proper techniques, even the soft parts and sometimes details of the internal structure can be recognized. Famous formations of this kind are the Middle Cambrian Burgess shale of British Columbia, where the ventral appendages appear as flattened, extremely detailed impressions; the Middle Ordovician Utica Shale of New York; and the Lower Devonian Hunsriick Slate of Germany. In the fossils from the latter two deposits, the trilobite soft parts are replaced by fine crystals of iron pyrite, and this fortunate feature lends itself to detection by visual observation and, more effectively, by radiography. Trilobites with preserved appendages have occasionally also been found in very finegrained limestones. From the above occurrences, the ventral appendages of not more than twenty species of trilobites have been described in several famous studies by Walcott, Raymond, Stormer, and others. These studies are comprehensively summarized in the Treatise (Moore 1959). A most unusual and complete description of the ventral anatomy of the larvae of agnostids from the Upper Cambrian of Sweden (Miiller and Walossek 1987) has given a glimpse of the extraordinary complexity of these reputedly primitive creatures. In recent years, improved X-ray techniques have yielded much new information on the anatomy of trilobite's soft parts, in particular from studies by J. Cisne (1973) on the trilobites from the Utica Shale and by Stiirmet and Bergstrom (1973) on the trilobites from the Hunsriick Slate. I had the fortune of working in close collaboration with Dr. Cisne when he was preparing his Ph.D. dissertation on the anatomy 18

The Trilobita

of Triarthrus eatoni (Hall), the most abundant of the trilobite species in the Utica Shale. The material at the disposal of Dr. Cisne was the exceptional collection of specimens assembled and prepared by C. E. Beecher toward the end of the last century and now belonging to the American Museum of Natural History, Field Museum of Natural History, Harvard Museum of Comparative Zoology, and the Yale Peabody Museum. The trilobites originate from a thin layer, about ten millimiters thick, in the black Utica Shale. Beecher quarried this layer extensively, collecting some seven hundred specimens, and exposed the appendages of some sixty specimens by gently rubbing away the matrix with pencil erasers. Several trilobites were further prepared by Dr. Cisne to make them suitable for soft X-ray examination, which was carried out by taking stereoscopic radiographs. In addition to furthering a better understanding of the already known appendages, Dr. Cisne's study revealed the presence of previously unsuspected details of the internal anatomy of the trilobites. As it turns out, the entire digestive tract of these specimens was often preserved, as well as muscle fibers and other surprising features of the trilobite's body. Several of Dr. Cisne's original radiographs were kindly loaned to the author for printing and are presented in this book (see plates 9, 11, 127a, 128). Furthermore, several specimens were selected and loaned to enable me to experiment with the optical techniques most suitable to enhance the visual contrast of the exposed appendages. Some of the photographs in this section and later in the atlas constitute a unique record of Beecher's trilobites. Figure 3 presents a reconstruction of the dorsal and ventral side of Triarthrus eatoni (Hall) based on the observations by J. Cisne. It differs from previous reconstructions in several details, the most prominent being the number of cephalic appendages (three instead of four pairs) and the presence of an abdominal extension carrying the anal tract. The latter protrudes beyond the border of the pygidium, a feature that could also be interpreted as due to compressional shift of the soft ventral structures relative to the tergum, caused by sediment load. The basic feature of the ventral anatomy is the presence of a pair of biramous appendages

carried by each of the thoracic segments. The first pair is modified into two segmented antennae, which serve an obvious sensory function. The cephalic region also contains three pairs of slightly modified appendages, and the pygidium five pairs, while about eight pairs are carried by the terminal abdominal tract. In total, approximately thirty-one pairs of limbs can be recognized. The basic structure of the biramous appendages is shown in figure 4. It is, in many respects, similar to that found in certain Crustacea. The base portion of the limb carries two differentiated branches: a featherlike or gill-bearing outer branch, thepre-epipodite (like the crustacean exopod), and an inner branch or walking leg, the telopodite (like the crustacean endopod), composed of seven articulated limb segments. Short bristles or setae appear at various locations. The featherlike structure of the outer branch has been interpreted as representing gill-blades, which performed the respiratory function. This interpretation, however, is still controversial. Nevertheless, there is no doubt that the inner branches were constructed for crawling. The tracks left by the motion of the latter on the soft sea floor are preserved occasionally as fossils trails, called Cruziana. An example of such trails is shown in plate 7. The outer branches could have helped in swimming, their shape and arrangement suggesting a 'Venetian blind" oarlike stroke. Although the telopodites terminate into diminutive, clawlike processes, it is doubtful that these were of much use in gathering food. Trilobites like Triarthrus seemed to have fed on detritus or microorganisms. Their food, in the X-ray plates of J. Cisne, can occasionally be observed as clouds of finely particulate material squeezed out of the gut canal following burial compression. The mouth is thought to have been a small opening, posteriorly oriented, just at the tip of the hypostome, devoid of jaws or mandibles. It would be the terminal feature of the long enclosure defined by the limb bases. This suggests a so-called trunk-limb feeding mechanism. It involves the creation of a feeding current by a rhythmical motion of the limbs, in this case making the telopodite bases convey the food particles to the mouth through the ventral food groove. The bases of the head limbs were somewhat differentiated from the thoracic limbs,

perhaps serving some kind of weak masticatory function. Locomotion, feeding, and mastication were probably part of a mechanically related 'sequence of events produced by the movement of the limbs. Once ingested, the minute food particles would pass through an esophagus into the stomach, located beneath the frontal glabellar lobe, and then into the intestine, a long tube running through the axial region and terminating in the anal duct. Still considered as part of the digestive system in the agnostids and called genal caeca is a network of ramifications radiating away from the axial region into the genal area of the cephalon. This anatomical feature gives rise to the prosopon (functional ornamentation) observed on the dorsal cephalic surface of many trilobites as a fine mesh of radiating and ramifying ridges. It is now believed to represent part of the vascular or circulatory system (Bergstrom 1973a) and will be repeatedly noticeable in the atlas photographs. Another unusual feature of Beecher's trilobites is the apparent preservation of the muscle structure. As observed by Dr. Cisne, this consisted of a very efficiently engineered network of longitudinal, dorsoventral, horizontal, and limb muscles, which would ensure articulation, limb movement, and enrollment. Figure 4 shows only a few of such muscles in a transverse cross section of the thoracic region. Figure 5 shows portions of a longitudinal cross section, indicating, on the one hand, the complexity of the trilobite design but, on the other, its rationality. Many details of other organic functions have been revealed in Dr. Cisne's study. I will limit the description here, however, to a presentation of some of the material that led to so much insight into the anatomy of Triarthrus. In plate 8 we see the ventral side of a specimen of Triarthrus, as originally prepared by C. E. Beecher. The photograph has been taken with the specimen totally immersed in xylene (see Appendix B). The optical contact established by this liquid of high refractive index eliminates surface reflections and enables the achievement of maximum optical resolution. The hypostome, the antennae, and most of the successive appendages can be seen. The pyritized organic material appears white against the black matrix. Plates 9 and 10 3.2

Appendages

and

Internal

Anatomy

PLATE

7

T h e furrowed trails left b e h i n d by

Salter, was collected by Jan Bergstrom

trilobites crawling over m u d d y sea floor,

from L o w e r O r d o v i c i a n sandstones found

are sometimes preserved as trace fossils.

on Bell Island, C o n c e p t i o n Bay,

This

20

example

of

Cruziana semiplicata

T h e Trilobita

N e w f o u n d l a n d , ( x l . 4 , R L S coll.)

FIGURE

3

A m o d e r n reconstruction of

Triarthrus eatoni

(Hall),

c o m p l e t e d from partial drawings b y C i s n e ( 1 9 7 3 ) . Details of the ventral side are occasionally o m i t t e d to show u n d e r l y i n g structure.

FIGURE

4

Structure of the biramous appendages and cross-sectional view of the thoracic region

in

Triarthrus eatoni

(Hall).

C o m b i n e d assembly from separate reconstructions by Cisne ( 1 9 7 3 ) .

FIGURE

5

Longitudinal section through parts of the body of

Triarthrus eatoni

(Hall).

This

drawing, adapted from reconstructions by C i s n e ( 1 9 7 3 ) , shows details o f the muscle structure a n d of the intestinal duct.

3.2

Appendages

and

Internal

Anatomy

r

represent another specimen of the same trilobite as seen in, respectively, a radiograph by Dr. Cisne and a photograph by the author. It must be remembered that the radiographs were taken in stereo pairs. The stereoscopic observations of such pairs enables the distinguishing of features occurring throughout the depth of the specimen. Such features overlap in a single projection and make their interpretation more difficult. The appendages are clearly visible, however, and can be seen even when underlying the dorsal shield. Although the radiograph has very high resolution, the featherlike gillbearing branches are not as visible as in plate 10, which was obtained by immersing the specimen in a bath of xylene, as for plate 8. The photographic technique employed here is showing its advantages. The gill-branch setae are now clearly discernible together with the segmented structure of the preepipodite axis, which carries them. Plate 11 is a radiograph of another specimen. Here the digestive tract is visible, in particular in the vicinity of the anal termination. Many of the anatomical details mentioned previously can be made out from these examples, in particular with the aid of a magnifier. It should be remembered, however, that no individual specimen will show all the features exhibited in the reconstructions illustrated in figure 3. It is only through the examination of a large number of specimens that the overall structure can be visualized. Other examples of Triarthrus W\\\ be shown in the atlas section, where the taxonomic information concerning this trilobite will also be given. Much more is contained in Dr. Cisne's dissertation, in particular very important observations relating trilobites to the other groups of arthropods. The other important study referred to—that of Sturmer and Bergstrom (1973), was preceded by the spectacular observations by W. Sturmer (1970). The late Professor Sturmer was a physicist colleague and, as myself, also a trilobite enthusiast. In his work at Siemens AG, he had access to advanced X-ray and image processing equipment. After retirement, he owned his own X-ray station. On several occasions, in his jovial and exuberant vein, my friend gave colorful descriptions of how he searched for fossils in the field, by X-raying freshly quarried slabs brought into his X22

The Trilobita

ray van parked at the bottom of a famous slate quarry. This slate quarry in Hunsruck, West Germany, is the depository of a paleontological treasure. The slate is of Lower Devonian age, 1,000 meters thick, containing a pyritized fossil fauna retaining the record of the soft parts of a variety of marine animals. Complete trilobites are found in the shale, with all appendages preserved, much as in the Utica Shale of Rome, New York; sudden burial in mud, in a favorable chemical environment, may have been responsible for the exceptional preservation in both occurrences. By applying contrast enhancement techniques to his radiographs, Sturmer was able to give a new description of old fossils, uncovering surprising details that had escaped previous investigations. A classic example of these observations is provided by the comparison between an ordinary photograph of a specimen of Phacops (plate 12a) and its radiograph (plate 12b). The latter made news immediately (Sturmer 1970): gracefully floating with the eerie appearance of a luminescent deep-sea creature, Phacops sp. WS 295 acquired a new dimension in contrast with its previous image in stone. The structure of the appendages, the exitic spines, the digestive tract are sharply delineated; it should be noticed that were the radiograph of plate 12b that of a living trilobite, the contrast would not be as good as for the pyritized fossil. The mineralization of the soft parts in the latter is actually acting as a heavy element staining of the specimen, which enhances radiographic contrast, much as is done in angiography and other medical X-ray procedures. Another of Stunner's radiographs will be shown in the next section. From a small quarry in Vastergotland, Sweden, cut into Upper Cambrian rocks, came one of the most surprising paleontological discoveries of recent years. Among the myriad of shields of the diminutive agnostid Agnostus pisiformis (Wahlenberg), which populate limestone nodules found in the quarry, Miiller and Walossek (1987) were able to isolate larval stages, mostly enrolled, enclosing in full threedimensional relief all the ventral soft parts of the animal, including integument, limbs, setae and hairs. Due to phosphatic replacement, these delicate organs could be cleaned of the embedding matrix by etching with weak acids.

PLATE

8

V e n t r a l view of a

specimen of Triarthrus

eatoni

(Hall), whose

pyritized appendages h a v e been exposed after painstaking p r e p a r a t i o n by C. E. Beecher ( x 4 . 7 ) . F r o m "Beecher's trilobite bed," Utica Shale, Upper Ordovician, Rome, N e w York. ( Y P M 2 1 9 , loaned through courtesy o f Dr. J . Cisne.) Photograph o b t a i n e d while the specimen was totally immersed in xylene.

3.2

Appendages and Internal Anatomy

23

PLATE

9

Radiograph of an adult specimen o f

Triarthrus

eatoni

(Hall) ( x 4 . 2 ) . ( Y P M 2 2 8 , from a stereo pair of radiographs by J . Cisne.) T h e b i r a m o u s appendages can be seen protruding from and underlying the dorsal shield. T h e pyritized soft parts of the trilobite, as well as the carapace, are m o r e opaque to X-rays than the e m b e d d i n g shale, thus providing high contrast. Details of other soft parts, replaced by fine granules of iron pyrite, yield visible contrast.

24

The Trilobita

PLATE

10

T h e same specimen as in plate 9, this time immersed in xylene. Fine surface details of t h e structure of the appendages are revealed by this technique.

3.2

Appendages and Internal Anatomy

25

PLATE

11

Another X-ray view of a c o m p l e t e l y preserved

specimen of Triarthrus

eatoni, as in the preceding plate (x7.4). (YPM

28253,

radiograph b y J . Cisne.) These pyritized trilobites are among the most striking fossil records of extinct life ever recaptured.

26

T h e Trilobita

PLATE

12

(a) Pyritized specimen of

Phacops

sp.

from the Lower

D e v o n i a n Hunsriick Slate, s h o w i n g the appendages, exposed by mechanical preparation ( x 2 . 1 ) ; (b) X-ray p h o t o g r a p h o f the same, b y W . S t u r m e r ( W S 2 9 5 ) . Both photographs were contributed by W. Sturmer.

3.2

Appendages

and

Internal

Anatomy

11

All observations had to be carried out with a scanning electron microscope, in view of the minuteness of the complex structures entirely contained within the submillimeter-sized larval shields. From this painstaking body of work, what emerged is a truly astounding reconstruction of the ventral view of this creature, reproduced in figure 6a, together with a sketch of its embryonic-looking dorsal appearance (fig. 6b).

FIGURE (a)

6

Reconstruction of the

ventral anatomy o f Agnostus pisiformis (Wahlenbetg), after scanning electron microscope observations of unusually well preserved enrolled larvae, (b) Dorsal view of adult

individual

(holaspid) of the same agnostid, typically only a few millimeters long, (c) Side view of enrolled individual.

(From Miiller

and Walossek permission

1 9 8 7 , by

of

Universitetsforlaget, Blinolern,

The

Trilobita

Oslo.)

Although substantially different from the structure of the multisegmented, or polymeric, trilobites, the agnostids have customarily been regarded as trilobites. However, the new detailed information on the structure of agnostid appendages that is summarized in the reconstruction of figure 6 reveals several features that are shared with primitive crustaceans, raising questions as to whether the agnostids are indeed trilobites at all. After the beautifully illustrated report by Miiller and Walossek appeared, I remembered collecting agnostids from the same area of Vastergotland, on a very hot summer day, accompanied by my friend and mentor Jan Bergstrom. It was a pleasant surprise to rediscover some of this material, and the result is shown in plate 13. The accomplishment of the above authors in revealing the inner structures encased in these miniature shields appears most impressive indeed. In order to conclude this section on a light note, plate 14 shows what polymeric trilobites actually may have looked like to the casual observer. The two creatures shown, one in the normal crawling posture, the other helplessly overturned, were photographed by the author on the patio of a beautiful resort in the mountains overlooking Oslo, where the International Conference on Trilobites was held in July, 1973. We are dealing here with reconstructions of Olenoides serratus (Rominger), prepared at the Paleontological Museum, University of Oslo, and displayed for the enjoyment of the convening paleontologists. In spite of their appearance in the plate, trilobites were not light-emitting animals. The picture is simply a print from a color slide and is therefore what is usually regarded as a negative. These trilobites possessed another pair of sensory appendages, posteriorly located and called cerci. They seemed to enjoy the midsummer Norwegian sun.

3.3

The Eyes of Trilobites

Not only the trilobites developed highly organized visual organs, but some of the recently discovered properties of trilobite's eye lenses represent an all-time feat of function

optimization. We are confronted here with a very successful scheme of eye structure: the composite or compound eye, made of arrays of separate optical elements, the ommatidia, pointing in slightly diverging directions and each performing an identical function. A network of photoreceptors and neurons translates the optical stimuli into an image perception. Evidence of the success of such a scheme is widespread experience, since the eyes of insects and crustaceans, in fact of most arthopods, still follow a design closely related to that developed by trilobites.

3.3.1

The Compound Eye

In modern arthropods the structural unit (ommatidium) is made of a sequence of functional subunits (see fig. 7). Facing the outside world is the dioptric apparatus consisting of a corneal lens in optical contact with a crystalline cone. Tiny images from a narrow field of view appear at the tips of the cones. Proceeding toward the interior of the eye beyond the cone there are two types of sttuctures. In the so-called apposition eyes (mostly found in diurnal insects, Crustacea, etc.), the photoreceptor, or rhabdom, is long and attached directly to the tip of the crystalline cone. On the other hand, in the superposition eye (found in nocturnal forms like moths, fireflies, etc.) a crystalline fiber, like a light guide, intervenes between the tip of the cone and a shorter rhabdom. A dark pigment may fill the space between cones in the apposition eyes, while in the superposition eye the pigment layer can migrate to surround either the cones or the crystalline fibers, when the eye is dark- or light-adapted respectively. Furthermore, the superposition eyes are generally constructed with very regular radial symmetry. This feature and the pigment migration in the dark-adapted eyes have been interpreted as enabling occurrence of collective phenomena, such as superposition of images or diffraction patterns due to adjacent ommatidia. The interpretation of the visual process in the compound eyes, however, has been the subject of great controversy since the pioneering work of Miiller (1826), who introduced the "mosaic" theory, according to which the compound eye is regarded as an assemblage of directional

3.3

The Eyes of Trilobites

29

PLATE

13

Disarticulated exuviae o f

pisiformis

(Wahlenberg),

Agrwstus

from the Upper

C a m b r i a n o f Vastergotland, Sweden. (x6.8, R L S coll.)

30

The Trilobita

PLATE

14

Dorsal and ventral view of reconstructions

of

Olenoides serratus

(Rominger), a M i d d l e C a m b r i a n trilobite. T h e models were prepared at the Paleontological M u s e u m , University of Oslo. Ptint from a color slide.

3.3

The Eyes of Trilobites

31

units, each yielding a point element in a mosaiclike reconstruction. Exner (1891) further developed the mosaic theory in his classic study of faceted eyes and introduced the distinction between apposition and superposition eyes. However, most of Exner's models of image formation by the dioptric system of the ommatidia turned out to be wrong. For example, Exner assumed that the crystalline cones in Limulus would form images at their tips due to a radial variation of the refractive index (cylinder lens). This is now known not to be true; in fact, the same images can be obtained from a homogenous scaled-up replica of the Limulus cone made of lucite when it is immersed in water (LeviSetti, Park, and Winston 1973).' Furthermore, Exner ignored the role of the crystalline tracts between cone and rhabdom, present in the so-called superposition eyes. What seems to emerge from modern research is that the term superposition is a misnomer (Horridge 1969), since no superposition of sharp images is ever observed in the plane of the photoreceptors when the crystalline tracts are in place. After more than eighty years of research, the only significant distinction found in the types of arthropod eyes is that there are eyes with or without crystalline threads, and, correspondingly, with short or long rhabdoms. What has become increasingly apparent is the role of the neurophysiological apparatus in modifying the type of response which could be inferred from purely optical considerations. The lateral inhibitory interaction can alter the effects of image overlap between neighboring ommatidia and sharpen contrast. (For a summary, see, for example, Hartline 1969). A functional distinction between the two types of eyes discussed above is still obscure but may not be as fundamental as originally thought by Exner.

' T h e shape o f each ommatidium o f Limulus has been found to be optimized for the maximum collection of light incident within a field of view of aperture angle ± 1 9 ° from the axis. T h e relevant phenomenon taking place in such a device is total internal reflection at the interface between the corneal medium, or refractive index n = 1 . 5 3 , and the fluid external to the cone, of refractive index n = 1.35.

32

The Trilobita

3.3.2

The Compound Eye in Trilobites

Although, with rare exceptions, we have no knowledge of the interna] structure of the eyes of trilobites, the fossil record yields astounding evidence of the dioptric apparatus, the outermost region of the ommatidia. This is due once again to the fact that only this portion was sufficiently mineralized to remain preserved in the fossilization process. In favorable circumstances, the eye lenses survived intact their long burial: they were already made of calcite crystals in the living animals! Not all trilobites possessed eyes—some, in fact, were blind; but others had enormous eyes which would take up most of the cephalic surface. Often the eyes were shaped in turretlike fashion, and their combined visual field could cover the animal's entire surroundings. Trilobite eyes, as we shall see, are revealing indicators of the habits of their carriers. Lacking the complete structure of the ommatidia, we cannot, of course, draw an exact analogy between the compound eyes of trilobites and those of modern arthropods. However, the external appearance and the lensar arrangement are often suggestive of a very close correspondence. If provided with the present knowledge, Exner would probably recognize apposition and superposition eyes in trilobites. Lindstrom (1901) instead recognized two different kinds of trilobites' eyes: a truly compound eye and an aggregate eye. The compound, or holochroal, eye is characterized by a close packing of the lens units, the entire visual surface being covered by a continuous pellucid membrane, the cornea. The lenslets vary in shape from thin biconvex to elongated hexagonal prisms. A different structure is presented by the aggregate, or schizochroal, eye. Here the lenses are separately encased and positioned by a cylindrical mounting, the sclera, and each lens is covered by its own cornea. The holochroal eye with thin biconvex lenses is thought to be the ancestral form (Clarkson 1975), already well developed in the late Cambrian, while the holochroal eyes with thick prismatic lenses and the schizochroal eyes appear in post-Cambrian times. Any close similarities between trilobites' and modern arthropods' eyes is more likely to exist

FIGURE

7

S t r u c t u r e of the o m m a t i d i a in the c o m p o u n d eye o f m o d e r n a r t h r o p o d s . Parts (a) and (b) are schematic sections through firefly.

the o m m a t i d i a o f the Pigment

migration

distinguishes the light-adapted c o n d i t i o n (a) from the darkadapted c o n d i t i o n (adapted

from

(b)

Horridgc

1 9 6 9 ) . Part (c) represents the structure o f the o m m a t i d i a o f the ant (adapted from Snodgrass

1 9 5 2 ) . The two

types of eye are distinguished by the ptesence of short a n d long thabdoms

in the ancestral form of holochroal eye, where the two branches of arthropod evolution were phylogenetically closer. As we shall see, there are in fact forms in the holochroal eye that suggest a superposition-type organization. On the other hand, there is no reason to suspect that the appositiontype was not already present in trilobites. The schizochroal eye is externally organized very much in the same way as the latter. The Phacopida, however, the principal possessors of the schizochroal eyes, have evolved a sophisticated lens structure apt to correct the optical defects of thick lenses. Identical structures are not known to exist in modern arthropods, perhaps since they developed when trilobites were already a well-separated stock. However, the trilobite eye studies described in this section prompted recognition

respectively.

that similar morphologies and Structures observed among the dioptric apparatuses of several living arthropods serve an analogous function, a testimonial of convergent evolution aimed at satisfying a seemingly universal striving toward optimization of vision. Most of the photographs in this section originate from negatives kindly loaned by Dr. E. N. K. Clarkson of the Grant Institute of Geology, University of Edinburgh, Scotland, who has in recent years carried out comprehensive studies of the visual apparatus of trilobites. The author has extensively collaborated with Dr. Clarkson on the study of the optical functions of several types of lens structures, and some of the discoveries emerging from this work will be described here.

3.3

The Eyes

of Trilobites

33

3.3.3

Holochroal Eyes

A schematic view of the range of optical structures in this type of eye is shown in figure 8, adapted from Lindstrom (1901). In addition to the basic hexagonal prism design, other lens contours were present, including square prisms. The optical behavior of these elements, while seemingly obvious in the simple biconvex lenses of Sphaerophthalmus, conceals a surprising feat, which will become apparent in the case of the long prisms of Asaphus and Illaenus in particular. Such prisms were made of single calcite crystals (Clarkson 1973a), and, to offset the strong birefringence of calcite, the crystals were oriented so that the optic axis always pointed in a direction normal to the visual surface. Only along this axis, in fact, does calcite behave as an isotropic medium. A precise understanding of the function of this unusual apparatus—whether as a light guide or as a focusing device—is still lacking. Furthermore, it is also not known whether or not crystalline cone and fiber optics accompanied the lens that is preserved. Conceivably, the long prismatic bodies could have encompassed both functions of the lenscrystalline cone assembly. Possible schemes of vision based

on this kind of eye structure will be discussed further. The number of individual optical elements in the holochroal eye could vary from approximately one hundred to more than fifteen thousand in a single eye, a range not very different from that found in insects. The actual size of the eye, however, often exceeds that of modern arthropods. Perhaps nothing better than the photographic record can convey a feeling for these structures and, at the same time, show the overall shape of the eye and its visual field. Plates 15 and 16 show scanning electron microscope (SEM) pictures (Clarkson 1973b) of some of the ancestral holochroal eyes from the olenid trilobites of the Late Cambrian period of Sweden. In plate 15, the eye of Ctenopyge (Mesoctenopyge) tumida is seen partly covered by the corneal membrane. Where this is still present, the swelling caused by the presence of the underlying lenslets is apparent. Where the membrane is removed, only the lensar pits remain of the visual surface. The thin lens profile can be made out by comparing the two regions. Plate 16 shows an intact eye of Sphaerophthalmus alatus. The external similarities of these primitive eye forms to those of some modern insects (for example, the ant) is quite remarkable. Plate 17 contains two

FIGURE

8

Cross-sectional

and

frontal

views of the visual surface in several holochroal eyes of

(a) Sphaerophthalmus

trilobites.

alatus

(Boeck), U . C a m b r i a n ,

Sweden

(x80);

(b) Cyrtometopus (Dalman),

clavifrons

Ordovician,

S w e d e n ( x 5 0 ) ; (c)

chiron

Holm,

Illaenus

Ordovician,

Sweden, ( x 5 0 ) . Adapted from Lindstrom

34

The

Trilobita

1901.

PLATE

15

Scanning electron microscope view of the

eye of Ctenopyge (Mesoctenopyge) tumida Westergard (x 1 1 7 ) . Upper Cambrian, S w e d e n . (Negative loaned by E. N. K. Clarkson; C l a r k s o n 1 9 7 3 b . ) T h e visual surface is partially covered by the corneal membrane.

PLATE

16

S E M view of the eye of a y o u n g specimen

of

Sphaerophthalmus alatus

(Boeck) ( x 2 0 3 ) , U p p e t C a m b r i a n , Sweden. (Negative loaned by E. N. K. Clarkson; Clarkson 1 9 7 3 b . ) T h e corneal m e m b r a n e covers the entite visual surface; however, the swellings due to the underlying lenslets are clearly visible.

3.3

The Eyes of Trilobites

3 5

views of the eye of Scutellum (Paralejurus) campaniferum (Beyrich), a Devonian trilobite from Bohemia. The toroidal shape of the combined surface of both eyes could evidently span a visual field close to 4lt radians. Here the corneal membrane is missing and the remarkable packing arrangement of the lens elements is clearly exposed. On a small scale, the lenses are arranged in a hexagonal pattern, much like the cells of a beehive. On a broader scale, a pattern of intersecting logarithmic spirals emerges, much like the arrangement of florets in a giant sunflower. The properties of the logarithmic (or equiangular) spiral have fascinated naturalists and poets for centuries (D'Arcy-Thompson 1942). It arises in nature whenever growth (linear expansion) combines with the necessity of preserving circular forms (rotational symmetry). The growth of the trilobite eye, most noticeably of the holochroal variety, was dominated by this theme, as Clarkson (1975) has pointed out. Just as interesting as the regularities of these spiraling patterns are the imperfections and mistakes in the packing program and the steps that have been taken to correct them. Various dislocations in the pattern have been repaired while still preserving the major symmetries. In plate 18 we see an SEM picture of a portion of the eye of Paralejurus brongniarti (Barrande), also from the Devonian period of Bohemia. Although the pattern arrangement is hexagonal, the exposed terminations of the lens elements appear quite spherical. Plates 19a and 19b show different views of the eyes of Pricyclopyge binodosa (Salter), from the Ordovician period of Bohemia. The preservation of these remarkable eyes is peculiar since only the ommatidial lens framework is preserved as an empty dome-shaped beehive. This indicates that between adjacent units there might have been a wall, which in this case is the only preserved component. The cross section of the lens assembly is exposed in cuts and holes, thus giving a measure of its thickness. This ttilobite had immense eyes in relation to the size of its body. The eyes were placed on the sides of the cephalon, so as to extend to the ventral region, and would almost touch each other in front. In related forms, the eyes actually merged into one uninterrupted visual surface which looked like a 36

The Trilobita

panoramic dome. The arrangement of the lenses in the eye shown in plate 20 is unusual. We see two generative growth zones, each of them initiating a spiraling pattern in opposite directions. The two diverging patterns come together along a straight line in the central region of the eye. Finally, plates 21a and 21b show two SEM views of a cross section through the visual surface of Asaphus raniceps Dalman (Clarkson 1973a). The words of Lindstrom (1901) in describing this structure are most appropriate here: the lenses "are columnar prisms, like the pillars of basalt." These are the structures made of oriented calcite mentioned in connection with the description of figure 8. As noted by Clarkson, the symmetrically radiating arrangement of the ommatidial prisms in this trilobite reminds one of the arrangement of the superposition eye of modern arthropods. For illustration of this point, figure 9 shows a cross section through the eye of Asaphus compared with that of a night moth, Deilephila elpenor (in the light adapted condition, in order to show the crystalline cones) (Hoglund 1966). The actual photographs of a thin film section through the Asaphus'eye is shown in plate 22 and a magnified detail of the array of hexagonal prisms that make up its visual surface is shown in plate 23. Each element is terminated by an essentially flat surface on the outside of the eye and by a convex spherical surface on the inside. These are disproportionally thick plano-convex lenses. As I mentioned earlier, the lens elements of trilobites' eyes were made of single-oriented calcite crystals. This is spectacularly demonstrated in plate 23, where the single-crystal nature and crystal orientation of each element are clearly revealed by the array of cleavage planes, which cross the prisms at a constant angle to the axis of each element. From the symmetry of the cleavage pattern, it can be deduced that the optic axis (the f-axis) of the calcite crystals is aligned in all cases with the lens element axis. The shape and proportions of the globular visual surface of the Asaphus' eye give us some clues about the inner workings of the holochroal eye. Judging from the profile and composition of the lens elements, and assuming that the eye cavity was filled with a medium that had refractive index not very different from that of water, one can conclude that each lens element must

PLATE

17

T w o views o f the same

eye of Scutellum (Paralejurus) campaniferum (Beyrich) ( x l 5 ) . (Negative loaned by E. N. K. C l a t k s o n . ) T h i s D e v o n i a n trilobite from B o h e m i a could cover an a l m o s t spherical

3.3

The Eyes of Trilobites

visual field.

37

have formed sharp primary images of a portion of the trilobite's surroundings about 0.2 millimeters below the inner vertex of each lens (a distance equal to the thickness of the lens). This presents a puzzle. For if the photoreceptors were this close to the surface of the eye, where they could make use of primary images, why was the overall structure of the eye so deep—as if it had been constructed to accommodate long rhabdomes? It may be that the photoreceptors were in fact placed further back in the eye than the evidence we have would suggest. Certainly, the possibility that collective diffraction phenomena might result from the array of primary images would have been a good reason for placing the photoreceptors further away from the visual surface. ( This was, in fact, recently discovered to be a relevant factor in the vision of insects, as reviewed by Burtt and Catton 1966.) Even though the principal Fourier image would appear much further behind the eye in this case, intermediate images could still be probed at several locations within the eye cavity, to the trilobite's great advantage. Unfortunately, we have reached at this point an impasse in our investigation. The marvelous optical interface at the surface of the holochroal eye is all that is left to us. We see nothing meaningful beneath it—nothing to either confirm or deny our suppositions about internal structures. PLATE

18

S E M picture o f portion o f the visual

surface (Barrande)

of Paralejurus

brongniarti

( x l 0 8 ) . D e v o n i a n , Bohemia.

(Negative loaned by E. N. K. Clarkson.)

38

The

Trilobita

PLATE

19

F r o n t (a) and side (b) view

of the eye of Pricyclopyge binodosa (Salter) ( x l 3 ) . Ordovician,

Bohemia.

(Negative loaned from E. N. K. C l a r k s o n . ) In this unusual process of fossilization, o n l y the f r a m e w o r k of the visual surface is preserved.

3.3

The

Eyes

of Trilobites

39

PLATE

20

T o p view o f the eye o f a n o t h e t specimen

of

Pricyclopyge binodosa

(Salter), as in plate 19 ( x l 8 ) .

40

The Ttilobita

PLATE

21

S E M views o f the prismatic structure o f the visual surface o f

Asaphus raniceps

D a l m a n , a L o w e r O r d o v i c i a n trilobite from N o r t h e r n O l a n d , S w e d e n (a, x l 5 0 ; b, x 3 3 0 ) . (Negative loaned by E. N. K. Clarkson; C l a r k s o n

3.3

1973b.)

The Eyes of Trilobites

41

FIGURE

9

(a) Cross-sectional view of the superposition eye of the night m o t h , in the lightadapted condition (adapted from Hoglund 1 9 6 5 ) . (b) Cross-sectional view of the

holochroal eye of Asaphus (adapted from Clarkson 1 9 7 3 b ) (see also plates 22 and 2 3 ) . The ommatidial arrangement was probably similar in the two cases.

PLATE

22

Panoramic view of an entire

eye section of Asaphus

raniceps

D a l m a n , from a thin

polished section observed under the optical microscope ( x l 6 , section loaned by E. N. K. Clarkson, G r . I. 5 5 1 2 ) .

42

The Trilobita

PLATE

23

G r e a t l y magnified detail of

the Asaphus eye section of the preceding plate, observed in dark field illumination. T h e calcite crystal nature and orientation of each element is clearly revealed by the array of cleavage planes, w h i c h can be seen crossing the prisms at a constant angle to the axis of each element. T h e optic axis of the calcite crystals is aligned in all cases w i t h the lens element axis. A l o n g such an axis, the ordinarily birefringent calcite crystal behaves as an isotropic transparent material,

thus

avoiding double vision ( x 4 6 0 ) .

33.4

Schizochroal Eyes

This type of eye, which probably evolved from the holochroal eye, is a visual system quite different from any other eye that has ever appeared in the animal kingdom. In the structure of the schizochroal eye, we see trilobites at the peak of their functional creativity, taking advantage of the fundamental laws of geometrical optics in a direct and most efficient way. The lenses in the schizochroal eye are generally larger (0.2 to 0.7 millimeters), and less numerous (from a few to a few hundred) than the lenses in the holochroal eye. Schizochroal eyes are turretlike, in the shape of truncated cones, and the lens arrangement is most orderly. Logarithmic spirals can still be seen, but the dominant theme is the dorso-ventral file—a rectilinear array of lenses running from

the top to the bottom of the visual surface (see plates 24— 33). The evolution of the schizochroal eye has been attributed by Clarkson (1971, 1975) to a phenomenon called paedomorphosis, by which some characteristics that were present in juvenile forms of some trilobites with holochroal eyes became the dominant characteristics in adults of descendant species. We find this type of eye in only one group of trilobites: the Suborder Phacopina, which contains the Superfamilies Dalmanitacea and Phacopacea—the dalmanitids and phacopids. This group lived for a period of about 150 million years during the Ordovician, Silurian, and Devonian periods. Similarities with the schizochroal eye are found in the eyes of the Cambrian Eodiscina (Jell 1975; Zhang and Clarkson 1990).

3.3

The

Eyes

of Trilobites

43

Plate 24 shows one of the earlier forms of the schizochroal eye and the trilobite that carried it: Pterygometopus brongniarti (Reed), a dalmanitid trilobite from the Ordovician of Scotland. In this particular specimen, the visual surface of the prominent eyes is made up of empty alveoli in which the lenses were once encased. A mold taken from the external impression of the same eye, which was left in the limestone matrix when the trilobite was removed, shows that the lenses were quite convex and closely packed (plate 25). A later dalmanitid form is that of Odontochile hausmanni (Brongniart), shown in plate 26. The lenses here are well separated, and the growth mechanism of the visual surface is plainly illustrated by lenses which are in the process of being released from the generative zone at the base of the eye. In the phacopid line, one of the earliest forms of schizochroal eye is seen in Acernaspis sufferta (Lamont), from the Lower Silurian of Scotland, shown in plate 27. The lenses here are more compressed together, giving them an almost hexagonal outline. This eye form eventually led to the perfected eye of the Silurian Eophacops trapeziceps (Barrande), illustrated in plate 28, and of the Devonian Phacops latifrons (Bronn), seen in plate 29. In these late species, the characteristics of the schizochroal eye are exaggerated: the eye contains relatively few lens elements that are rather widely spaced, and each element is admirably set in a thick, sometimes swollen sclera. It is hard to refrain from presenting additional examples of these superbly architectured eyes. The eyes of Phacops rana crassituberculata Stumm from the Devonian Silica Shale of Ohio are shown in plates 30 and 31. These eyes show clearly a band of "sensory fossettes" below the visual surface. The lenses are quite deeply set inside the sclera. A variant of this trilobite, also from the Silica Shale, is Phacops rana milleri Stewart, whose eye is shown in plate 32. It contains more lenses than the previous one (see, for example, Eldredge 1972). The large trilobite Phacops (Drotops) megalomanicus, from the Devonian of Morocco has in recent years pervaded the shelves of every rock shop throughout the world. Its morphology is quite similar to that of, for example, Phacops rana crassituberculata, although in much expanded form, reaching a length of up

44

T h e Trilobita

to 15 cm, and its carapace more prominently tuberculated. Its eye is proportionally magnified, yet the lens size is not, as can be appreciated from plate 33. The increased spacing between the lenses left room for the sclera to develop a prominent hexagonal set of tubercula. In most phacopid eyes, the field of view of each eye could span 180 degrees longitudinally and a strip of ten to twenty degrees latitudinally. The schizochroal eye with the smallest number of lenses is shown in the SEM photograph on plate 34. It is the eye of Denckmannites volborthi (Barrande) of the Devonian of Bohemia. In the face of such a layout it becomes difficult to accept the notion that the optical interface of the schizochroal eye operated by fragmenting images of the outer environment into a mosaic of relatively few point elements. If each one of the relatively small number of lenses provided just one point element of the picture, such a mosaic would be very crude indeed. There is evidence that the schizochroal eye was much more efficient than that, and the most compelling part of this evidence is the astonishing structure of the schizochroal eye lens itself.

3.3.5

Trilobite Eyes and the Optics of Descartes and Huygens

The biconvex lenses of phacopid trilobites are made up of doublet structures that were constructed for an unmistakable purpose: to correct for the large spherical distortion (aberration) of simple thick lenses. I reached this rather surprising conclusion after examining the large amount of evidence on lens morphology that had been accumulated by Euan N. K. Clarkson (1968) during his systematic investigation of trilobite eyes. A preliminary account of this discovery was contained in the first edition of this book (Levi-Setti 1975). Dr. Clarkson and I have since copublished a paper in Nature (Clarkson and Levi-Setti 1975) that describes this evidence and our interpretation of the extraordinary function of such doublet structures. When we humans construct optical elements, we sometimes cement together two lenses that have different refractive indices, as a means of correcting particular lens defects.

PLATE

24

O n e of the earlier examples of schizochroal eye, carried by the O t d o v i c i a n d a l m a n i t i d ttilobite

Pterygometopus

brongniarti

(Reed).

T h e eye lenses are missing in this specimen, but the framework of alveoli in which they were set has remained intact. (Collected at G i r v a n , Scotland, on a field trip guided by E. N. K. C l a r k s o n ) . (x8.3)

3.3

The Eyes of Trilobites

45

PLATE

25

Internal latex m o l d of the eye

of

Pterygometopus

brongniarti

(Reed), s h o w n in the preceding plate. T h e lenses are convex and arranged in vertical dorso-ventral files. (x!8)

46

T h e Trilobita

PLATE

26

A later form of schizochroal eye, that of

Odontochile hausmanni

(Brongniart),

process of being released from the generative zone at the base of the eye.

has clearly separated lens elements.

(Specimen from the Lower D e v o n i a n of

T h e growth mechanism of the visual

T e t i n , Bohemia. Loaned by E. N. K.

surface is illustrated by lenses in the

Clarkson.)

(x!9)

3.3

The Eyes of Trilobites

47

PLATE

27

T h e schizochroal eye of o n e of the earliest phacopid ttilobites,

Acernaspis (Eskaspis) sujferta (Lamont). The lenses are closely packed and almost hexagonal in shape. Latex mold from external impression, specimen from the Lower Silurian of the Pentland Hills, Scotland. (Collected on a field trip guided by E. N. K. C l a r k s o n , RLS coll.) ( x ! 5 )

48

The Trilobita

PLATE

28

Eophacops

trapeziceps

(Barrande) exemplifies later, perfected forms in the e v o l u t i o n of the schizochroal eye. T h e lens elements are rather w i d e l y spaced and p r o m i n e n t a b o v e the plane o f the sclera. ( S p e c i m e n from the Silurian o f B o h e m i a . Negative loaned by E. N. K. Clarkson.)

PLATE

(xl8)

29

T h e right eye of Phacops

latifrons

(Bronn),

from

the

Devonian of Germany. (Loaned by E. N. K. Clarkson.) ( x l 4 )

The

Eyes

of Trilobites

49

PLATE

30 Phacops crassituberculata

Left eye o f

rana

S t u m m , from the D e v o n i a n Silica Shale at Sylvania, O h i o ( x l 3 ) . (RLS coll.; now at F M N H . ) Specimen whitened with magnesium oxide. T h e lenses at the top of the visual surface appear incompletely developed.

50

T h e Trilobita

PLATE

31

Right eye of another

specimen of Phacops

rana

crassituberculata

S t u m m ( x ! 3 ) . (RLS coll.; n o w a t F M N H . ) Specimen whitened with m a g n e s i u m oxide. T h e lenses are deeply encased in the scleral surface.

3.3

The

Eyes of Trilobites

51

PLATE

32

T h e right eye of

Phacops rana milleri Stewart, from the D e v o n i a n Silica Shale o f Sylvania, O h i o ( x l 3 . 5 ) . (RLS coll.; n o w at F M N H . ) This trilobite, familiar to m a n y collectors, differs from the subspecies shown in plates 31 and 3 2 , mostly in the larger number of eye lenses and in the less tuberculate surface of the

52

palpebral lobe.

The Trilobita

PLATE

33

W e l l k n o w n to the trilobite collector are

m o u n t a i n s of M o r o c c o . Its eye, as shown

tubercula o c c u p y the increased spacing

the giant Devonian Phacops (Drotops)

in this example, is about twice as large

between the lenses in a regular

megalomanicus, (aff. Phacops rana crassituberculata), that are extracted

as that o f the typical in

staggering n u m b e r from the Anti-Atlas

Phacops rana

shown

in the preceding plates, yet its lenses,

hexagonal pattern that mimics the lensar a r r a n g e m e n t . ( R L S coll.) (x7.7)

deeply set, are smaller. P r o m i n e n t scleral

3.3

The Eyes of Trilobites

53

PLATE

34

S E M view of the eye of

Denkmannites (Barrande).

volborthi Devonian of

Bohemia ( x 6 5 ) . (Negative loaned by E. N. K. Clarkson.) This type of schizochroal eye is among those containing the smallest n u m b e r of lenses.

In fact, this optical doublet is a device so typically associated with human invention that its discovery in trilobites comes as something of a shock. The realization that trilobites developed and used such devices half a billion years ago makes the shock even greater. And a final discovery—that the refracting interface between the two lens elements in a trilobite's eye was designed in accordance with optical constructions worked out by Descartes and Huygens in the mid-seventeenth century—borders on sheer science fiction. Of course, the laws of physics existed prior to their discovery by man. And we shouldn't perhaps be too surprised that the drive to optimize biological function—one of the fundamental evolutionary forces in all biological organisms—caused trilobites to follow physical laws to the fullest possible extent in their development of visual systems. The real surprise should not be that they did construct eyes that work according to the laws of physics, but that they did it with such ingenuity. The basic lens designs recognized in the original studies by Clarkson are reproduced in figures 10b and l i b . These drawings schematize the eyes of two 54

The

Trilobita

dalmanitid trilobites: on the left, Crozonaspis struvei (Henry), from the Middle Ordovician of Brittany; and on the right, Dalmanitina socialis (Barrande), a Middle Ordovician species from Bohemia. The thick biconvex lens shape of both results from the matching of two basic parts: an upper lens unit and an intralensar bowl. A wavy interface separates (and unites) the two components. Long before trilobites were even recognized as ancient inhabitants of our planet, Descartes in his La Geometrie (1637) and Huygens in his Traite de la Lumiere (1690) both had derived the general shape that the second refracting surface of a lens should have to have in order to eliminate spherical aberration—a defect of simple lenses which causes a point object to be imaged as a blurred disk. To recreate an exciting moment of my literature search, it is still worthwhile to reproduce in figures 10a and 11a the pages of these treatises that contained such derivations. Since the two authors assumed somewhat different shapes for the first refracting surface, they reached somewhat different conclusions about the shape of the second (correcting) surfpce.

FIGURE

10

(a) C o n s t r u c t i o n of a lens free of spherical aberration, from C . H u y g e n s '

la Lumiere.

Traite' de

(b) Cross-sectional

view of the lenses in the eye of

Crozonaspis struvei

Henry,

an O r d o v i c i a n trilobite from Brittany (Clarkson

1968).

T h e intermediate surface is shaped accordingly to the prescription by H u y g e n s .

3.3

The

Eyes of Trilobites

55

FIGURE

11

(a) C o n s t r u c t i o n , similar to that of figure 1 0 , after

Descartes in La Geometric Here again, the shape of the second surface makes the lens free of spherical aberration. (b) Cross-sectional view of the lens structure in the eye

of Dalmanitina socialis (Barrande) (Clarkson 1 9 6 8 ) . T h e intermediate surface is shaped in remarkable accord with the design by Descartes.

56

The Trilobita

However, the physics involved is the same in both cases and consists of an application of Fermat's principle of least time to achieve the so-called stigmatic condition for an optical system: to obtain a point image of a point object, parallel beams of light must traverse the path between the object and the image in the same amount of time—either by traveling along a straight line connecting the two points (the shortest path and therefore the fastest route), or by traveling along any other off-axis path through the lens. In more modern language: all rays from a point source will converge to a point image if they traverse minimal and identical optical paths. A lens that satisfies this condition is free of spherical aberration and is called aplanatic. Its exit surface will depart from a spherical profile being one of a family of mathematical curves called Cartesian Ovals (described by a fourth-order equation in Cartesian coordinates). By comparing the shape of the aspheric lens exit surfaces constructed by Huygens and Descartes with the two lens structures identified by Clarkson (figures 10b and 1 lb), little doubt remains that trilobites utilized the properties of Cartesian Ovals more than 400 million years before the seventeenth-century masters discovered the principle. This is not all, however. A difference exists between the media for which the theoretical profiles were derived (a glass lens in air) and those involved in the trilobite's lens operation. We know in fact that the upper lens units of a trilobite's lenses were made of calcite oriented along the oaxis (refractive index n = 1.66), much as previously seen for the holochroal eye, and confirmed by Towe (1973) for the phacopid eye lenses, and that they operated in water (n = 1.33). If we attempt to trace light rays through such a system, of weaker converging power than the glass-air system, we find that the lens would not focus all rays to a point as expected. In fact, the more peripheral rays would diverge. Here the role of the intralensar bowl comes into play to restore the focusing property of the lens and the correcting function of the interface. The unknown refractive index of the intralensar bowl can be inferred from the construction shown in figure 12a. This illustrates the optical function of the doublet structure and also shows (on the right side of

the lens axis) what would happen if the lens were a single solid unit. In the latter case, spherical aberration would prevent the formation of sharp images and would also dilute the amount of light collected along the axis of the lens. An experiment with a large scale reconstruction of the doublet with an upper unit made of calcite with its r-axis oriented along the lens axis, combined with an intralensar bowl made of transparent plastic (« = 1.63) that is shaped as in Crozonaspis, does indeed produce the results that were envisioned by Descartes and Huygens (and the trilobite), as shown in figure 12b. One can speculate that a perfusion of a small amount of soft organic tissue into oriented calcite, the materials available to the trilobite, could have easily reduced the refractive index of the calcite to the value inferred for the intralensar bowl. The realization that trilobites made recourse to a doublet lens structure to achieve the goal of improving their vision left me with the uncanny feeling that, if needed, nature could have equally well developed other multi-element optical instruments that are touted as unique creations of human ingenuity. In our case, a doublet structure is added to the already sophisticated aspheric correcting interface. The design of the trilobite's eye lens could well qualify for a patent disclosure. Prior art would mention the Schmidt plate of modern telescopes, a Cartesian surface performing function similar to that of the wavy interface of the trilobite's eye lens. As I will discuss further below, there is a follow-up on the trilobite's eye story, dealing with the eye lenses of several modern arthropods, that makes my views much less facetious than they may seem at this point. There is even more than meets the eye in the trilobite's feat. Lenses such as described are clearly optimized in more than one way. The fact that the upper unit is made of calcite, with its r-axis lined up along the axis of the lens, accomplishes two things. First, the refractive index of calcite is highest for this orientation (n = 1.66, which maximizes the light gathering ability of the lens). Second, the double refraction effect of calcite is effectively eliminated by this configuration, at least for paraxial rays. As a result, such lenses succeed in concentrating all the collected light in a 3.3

The Eyes of Trilobites

57

FIGURE

12

(a) Ray tracing through the lens of

indicated in the text. This suggests that

calcite oriented with the c-axis along the

Crozonaspis struvei

the t w o lens elements were made of

lens axis, the intralensar bowl of

of the optical axis, the internal lens

oriented calcite and protein-rich material

transparent plastic with refractive index

structure is ignored. V e r y large spherical

respectively, (b) Experimental verification

n = 1 . 6 3 . T h e lens is immersed in milky

aberration ensues for any choice of the

of the operation of a large-scale doublet

water. A beam of parallel light incident on the lens from above is brought to a

Henry. O n

the right

refractive index of the lens (» = 1 . 6 6 for

lens structure modeled after the lens of

the construction indicated). W h e n the

Crozonaspis struvei

internal structure is taken into account,

and Levi-Setti 1 9 7 5 , reprinted by

Henry

Nature

(from

Clarkson

on the left-hand side of the axis,

permission from

correction of spherical aberration obtains

Copyright ©

for the combination of refractive indices

Limited). T h e upper lens unit is made of

58

The T r i l o b i t a

254: 663-67;

1 9 7 5 Macmillan Magazines

narrow focus by the corrected lens. T h e full width at half m a x i m u m of the light intensity depth distribution w o u l d be five times wider for a single uncorrected lens of the same overall external profile.

thin layer (approximately one lens-thickness below the vertex of the lens) where good images of the surroundings would be formed. The f: number (inverse of the relative aperture) of the lens of figure 12, operating in water, is - f: 1.1. Not bad at all, even by modern standards. And because of their size and short focal length, our wonder lenses give to the trilobite's eye a remarkable depth of field with no need for accommodation. Further analysis of the possible effects of the birefringence of calcite suggests that they were probably unimportant (Clarkson and Levi-Setti 1975). As to lens defects due to chromatic aberration, these also were considered unimportant, since even at moderate depths in sea water, the environment is essentially monochromatic. In the actual fossilized eye of phacopids, the evidence for the extraordinary doublet structure is occasionally obvious in polished sections of the lens surface. An example is provided by the series of images in plate 35, obtained from a polished thin section through the large eye of a Silurian Dalmanites specimen, prepared by Dr. Clarkson. Here dark field microscope illumination enhances the visibility of the doublet structure, and the shape of the correcting surface is intermediate between the two solutions previously mentioned. The characteristic calcite cleavage planes are clearly outlined in the upper unit of several of the lenses, again providing evidence of their single crystal nature and f-axis orientation. The intralensar bowl appears as a dark, cloudy region, suggesting that its chemical composition was different originally from that of the upper unit. Even when diagenetic alteration has rendered the upper unit opaque, the intralensar bowl can still be identified as a distinct region of the lens. In other cases, due to differential mineralization, the intralensar bowl splits from the upper lens unit. For example, the doughnut shape of the upper surface of the intralensar bowl is plainly visible in the SEM image shown in plate 36, where the upper lens units were not preserved in the fossilization process. The visual surface is, in this case, that of Zeliskella Lipeyrei (Bureau), from the Ordovician of the Crozon peninsula, France, which has a structure similar to that of Crozonaspis. Some remnants of the doublet structures are present in the holotype of Dalmanitespratteni Roy

(Roy 1933), a spectacular phacopid trilobite from the Devonian of Illinois. Front views of the right and left eye of this rare trilobite are shown in plates 37 and 38 respectively. Each eye contains more than 770 lenses, an absolute record for schizochroal eyes. Several of the lenses are split and show the intralensar bowl. In other cases, the upper lens unit is still in place, and in still others the entire doublet is missing and the cavity left is the lensar pit. The Huygens' type lens substructure has been established by Clarkson (1969) for the eye of Reedops sternbergi (Hawle and Corda), represented in plate 39. The evidence, however, is accessible only through sectioning. Two SEM views of the visual surface of this trilobite are shown in plates 40a and 40b. Why did the phacopid trilobite develop such a sophisticated optical system? Were the perfected images produced by the corrected lenses exploited in any way? Were there other advantages that favored the evolution and retention of these optimal lens structures? What we would like to hear, to appease our Darwinian upbringing, is that new visual structures were evolved in response to new environmental pressures as a means of survival. A factor that may be considered is that the correction of spherical aberration increases significantly the level of light intensity in the focal plane of the lens relative to that concentrated by a solid lens. From the experiment of figure 12b, we did estimate (Clarkson and Levi-Setti 1975) that the full width at half maximum of the light intensity distribution along the axis would be contained, for the trilobite's lens, within a layer - 20 Lum thick. This is to be compared with a width of - 100 Llm for an uncorrected, solid lens of the same profile, due to the spreading of the collected light along the lens axis caused by spherical aberration. The fivefold increase in the level of illumination at the focal plane could conceivably have exceeded the threshold level of neural response in a dimly lit environment, allowing the trilobite to see at some depth in the sea, at dusk, or in turbid water. And yet the lens arrangement and shape of the schizochroal eye raises doubts that a useful mosaic image could have been formed by this type of eye. The number of lenses is generally too small and the angular coverage of their fields of view too discontinuous 3.3

The Eyes of Trilobites

59

60

The Trilobita

^ PLATE

35

PLATE

(Facing page) T h e doublet

36

T h e d o u g h n u t shape o f the

structure of the dalmanitid

intralensar b o w l is clearly

lens is clearly visible in these

s h o w n in this scanning

thin sections from o n e

electron m i c r o g r a p h o f the

specimen of a Silurian

eye of Zeliskella lapeyrei

Dalmanites.

T h e upper lens

(Bureau), an early dalmanitid

unit consists of oriented

trilobite from the O r d o v i c i a n

calcite crystals, while the

of Rennes, France. In this

intralensar bowls are

specimen, o n l y the l o w e r lens

composed of calcite mixed

units have been preserved.

with remnants of organic

(Gr. I. 4 0 1 9 2 , loaned by E.

material. Various stages of

N. K. Clarkson, S E M

diagenetic alteration of the

micrograph by the author.)

calcite are visible, but n o n e

(xl 13)

of these obliterate the primary structure. M u c h as for the prisms o f the

Asaphus'

visual surface in plate 2 3 , the trace of the cleavage planes tell that the optic axis of the calcite crystals is along the lens axis. T h e Cartesian surface separating the t w o lens elements is recognizable in all cases s h o w n , as well as in some sixteen additional lenslets from the same eye, not shown here. (Polished section loaned by E. N. K. Clarkson.)

(xl08)

3.3

The Eyes

of Trilobites

61

PLATE

37

Frontal view of the

Dalmanites pratteni Roy, a left eye o f

D e v o n i a n trilobite from Illinois ( x l 3 ) . (Specimen loaned by Field Museum of Natural History, Chicago.)

62

T h e Trilobita

PLATE

38

Frontal view of the

right eye of Dalmanites

pratteni R o y , as in plate 3 7 . T h e lens structure described in the text is exposed in several instances, w h e r e the front lens-unit has split away. In s o m e elements, the entire lens appears intact, encased in the cylindrical sclera. A m o n g schizochroal eyes, the eyes of this trilobite exhibit the largest n u m b e r of lenses ever recorded.

3.3

The Eyes

of Trilobites

63

PLATE

39

T h e eye of Reedops Sternberg! (Hawle and C o r d a ) , a Devonian trilobite from Bohemia (x22). (Negative loaned by E. N. K. Clarkson; Clarkson

1969.)

Sectioning of the lenses of similar specimens by Clarkson has revealed an internal structure similar to that described in figure 1 0 (b).

64

T h e Trilobita

PLATE

40

(a) S E M view o f portion o f the

eye

of

Reedops

Sternberg!,

a s i n plate 3 9 ( x l 0 8 ) . (Negative loaned by E. N. K. C l a r k s o n . ) (b) A n o t h e r S E M view, at a glancing angle, of the same visual surface. (x216)

65

to form a detailed mosaic similar to that that we presume formed by the schizochroal eye and that of insects and crustaceans. Further questions come to mind: was the schizochroal eye a regressive trait? Was the improved lightgathering ability the only factor promoting the correction of lens aberrations? It is somewhat disappointing to conceive that the sharp and detailed images produced by each corrected lens were simply used as a trigger by the photoreceptors and their structure ignored by the "brain" of the trilobite. Questions such as these obviously stem from our ignorance of the visual receptor structure of the trilobite's eye. Claims of having detected crystalline fibers in the pyritized eye of phacopid trilobites, preserved in the Hunsruck Slate, have been advanced by W. Sturmer, based on his radiographic observations. If true, this knowledge would set to rest any notion that the schizochroal eye was anything more than what is found in the compound eye of groups of extants arthropods. Dr. Sturmer tried hard to convince his dissenting collaborator Jan Bergstrom, and then myself, of his interpretation. In the process, he provided me with the radiograph shown in plate 41, which shows a pattern of parallel filaments seemingly emerging from the visual surface and leading to points located in the posterior part of the cephalon. Additional examples of a similar nature (Sturmer and Bergstrom 1973) have been presented and interpreted as crystalline fiber lightguides by W. Sturmer, against J. Bergstrom's better judgment. As explained to me by the latter author, who had occasion to examine Sturmer's material, the observed structures do not belong to the eyes at all, but are gill branch setae, seen as a single file extending from the eye region to the pygidium. In some sense, this rejection is welcome. In fact, fiber optics coupled to phacopid lenses would seem to defeat the purpose of the sophisticated design of the latter, as already remarked above. Furthermore (Clarkson and Levi-Setti 1975), the hypothesis is advanced that schizochroal eyes may be regarded as an aggregate of individual eyes rather than the mosaic-forming device common to other arthropods. It is tempting to speculate that the schizochroal eye was a transition from the truly 66

The

Trilobita

compound eye scheme of the holochroal eye to the cameratype eye of more advanced life forms. Perhaps minuscule retinas in these schizochroal eyes could already analyze the images supplied by the state-of-the-art lenses they possessed. If this were so, one can speculate that image contrast may have been the factor prompting selection of ever better corrected lens systems to secure survival. In fact, the improved contrast resulting from the correction of spherical aberration could well have provided the advantage of a prompter recognition and response to impending dangers. Perhaps, in addition, mating may have proven more efficient with sharper images. The schizochroal eye disappeared toward the end of the Devonian, with the extinction of the phacopid trilobites. Was the evolutionary wonder represented by their eye lenses an unchallenged occurrence?

3.3.6

A Treasure Not Lost

Even if the genetic information of the perfected visual apparatus in the phacopid trilobite's eye became lost to further evolution within the phylum, the fundamental principles of physics that guided its development obviously survived. And, indeed, they guided other unrelated creatures to reproduce the mastery. From recent studies of vision in modern invertebrates, it has become apparent that the correction of spherical aberration following the precepts of Descartes and Huygens, as well as the concept of adopting a doublet structure for the dioptric apparatus, have not been the unique prerogative of the trilobites. The corneal thick lenses in the compound eye of the backswimmer Notonecta glauca, a predatory aquatic insect, have been shown (Schwind 1980) to consist of a doublet structure with an unmistakable bellshaped optical interface. Much as for the phacopid lens, the lower unit has a refractive index slightly lower than the upper unit, except that no calcite is involved in the lens composition, only organic material. Theoretical calculations and experimental determinations of the focusing properties of these lenses have confirmed that they are well corrected for spherical aberration (Horvath, 1989). Although we believe that the structures observed in the trilobite's lenses are real

PLATE

41

X-ray radiograph of a pyritized specimen of

Phacops

sp. from the Hunsriick Slate

c o m p o u n d eyes. A m o r e likely interpretation, discussed in the text,

o f the Lower D e v o n i a n o f G e r m a n y ( W S

assimilates the filamentary structures

2 6 1 7 , x 4 . 5 ) . T h e presence of a planar

w i t h those of the gill branches or

structure of filaments superposed to the

exopods, c o m m o n l y preserved in these

palpebral lobes was interpreted by W.

pyritized trilobites and often seen

Sttirmer as evidence of crystalline fibers

underlying

belonging to the o m m a t i d i a of the

c o n t r i b u t e d b y W . Sttirmer.

the carapace. Radiograph

3.3

The Eyes of Trilobites

67

(not, for example, due to diagenetic alteration of the calcite crystals), and that our interpretation of their function is sufficiently supported by our model, it is gratifying to find confirmation of our conjectures in a living system that can be studied without the need of assumptions. Since no direct connection can be seen between trilobites and Notonecta, it must be inferred that the similarity in the solutions to the problem of vision optimization was the result of convergent evolution. Another most intriguing two-component, corrected optical system is that found in the eyes of the scallop Pecten (Land 1965) and only recently brought to my attention (Horvath and Varju 1991). In this bivalve mollusc, fifty to sixty simple eyes are embedded in the pallium and appear as tiny, bright iridescent pearls. Their optical structure, sketched in figure 13, rivals in perfection and ingenuity that of the phacopid trilobite's lenses. They also consist of a compound, two-element structure. The upper unit is a soft lens, almost identical in shape to Huygens' solution, but mounted in inverted geometry, so that the bell-shaped interface is external. Facing the spherical, internal interface of this lens is a spherical mirror, the argentea, made out of thirty to forty layers of guanine crystals, interleaved with layers of cytoplasm. This multilayered structure acts as a highly reflecting, interferometric, quarter wavelength mirror. The image is formed on a retinal surface, located between the mirror and the correcting lens, that responds to a decrease in the level of illumination (the "off' signal). Another retinal surface, located beneath the former, responds to "on" signals only. The ensuing neural response imparts to the eye remarkable sensitivity to dimming of light levels and angular movement of the light-dark stimulus. In other words, the eye takes advantage of the image contrast, as surmised in our previous discussion of possible evolutionary advantages of correcting eye lens defects. Astonishing as this may seem, the two-element structure of the scallop's eye corresponds to the structure of the catadioptric telescope or Schmidt optical system. This compound lens system has an amazingly large angular acceptance, expressed by an f: number equal to 0.6.1 should mention that the Cartesian Ovals connection 68

The Trilobita

was not recognized in the earlier studies of the Pecten 's eye, although the function of the aspherical lens in correcting spherical aberration was fully documented. I also became aware of this preexisting evidence, after having already formulated my reflections, expressed earlier, concerning the significance of such complex designs found in naturally evolved living systems. Whatever repetition this may involve, I felt compelled to narrate how my premonition became eclipsed by reality. Indeed, a wide-angle imaging lens, inspired by the design of the Pectens lens system, has been incorporated into a miniaturized fiber optics endoscope, named the "Tube Peeper" (Greguss 1985). I indulged in a limited search, among living arthropods, for some evidence that the trilobite's visual treasure may not have been irretrievably lost. My quest dealt with an attempt to find whether any living marine invertebrates may have exploited calcite to construct their eye lenses. Although the search was unsuccessful from this standpoint, I enjoyed the opportunity of becoming aquainted with a group of marine Crustacea that evolved compound eyes and body morphologies quite similar to those of trilobites. These are the antarctic isopod Crustacea of the genus Serolis. I was alerted to their existence by a letter from David K. Bernhardt of the State University of New York at Albany, following publication of the first edition of this book. Mr. Bernhardt had located, in the local archives of the Albany Institute of History and Art, a nineteenth-century report announcing the discovery of a trilobitelike crustacean that may have possessed calcific eye-lenses. This letter triggered immediate follow-up on my part. The report (Eights 1833), that I promptly obtained from the Albany Institute, was written by James Eights, "Naturalist in the Exploring (Antarctic) Expedition of 1830," and Corresponding Member of the Institute. It described the discovery of a new "Crustaceous Animal" found on the shores of the South Shetland Islands that he named Brongniartia trilobitoides from its resemblance to trilobites and in honor of Alexander Brongniart, eminent paleontologist of the time who first developed the systematics of trilobites. The reading of Eights's report was most rewarding: he mentions that the eyes of his new find were "elevated

FIGURE

13

Schematics of the t w o - e l e m e n t structure in the single eyes f o u n d in the pallium o f the scallop

Pecten

(adapted from Land

1 9 6 5 ) . A lens of characteristic Huygensian shape corrects the otherwise aberrated image f o r m e d by a spherical mirror, the

argentea.

T h i s type o f

structure corresponds closely to that of the catadioptric telescope or S c h m i d t optical system. In the scallop, t w o retinal surfaces, located w h e r e the image is formed between the lens a n d the argentea, respond respectively to "on" and " o f f light signals.

and prominent: cornea oblong, lunulate, reticulate, composed of an infinite number of facets, distinctly visible to the naked eye: color blue, the superior surface covered by an irregular calcareous incrustation." The anatomy of this isopod was meticulously described by detailed ink drawings. The report also recreates in vivid color the misty, frigid, and yet sublime athmosphere and scenary of these antarctic shores being seen for the first time. It did not take anything more to fire up my curiosity and the desire to examine firsthand these blue-eyed little beasts. I poured over the report (Beddard 1884) on the Isopoda collected by the Challenger Antarctic Expedition of 1873-76, where I learned that Eights's "trilobite" is now called Serolis trilobitoides (Eights) and identified as an Isopod crustacean. Another report (Sheppard 1933) on the serolids collected by the Discovery Expedition (1925-32) lists some thirty-seven species of Serolis, all found in antarctic waters, some at great depths in the ocean. I then proceeded to borrow several Serolis specimens from the U.S. National Museum and alerted Dr. Clarkson that we could perhaps learn something about the habits of trilobites if we could observe the behaviour of live, look-alike Serolis. My search for calcite in the eyes of Serolis, carried out with radiographs and X-ray diffractometry, failed to detect crystal lenses. Calcite was found, but only external

to the eye, in the form of incrustations, as originally surmised by Eights. The internal structure of the eye of several serolids is described in the report by Beddard (1884). The dioptric apparatus contains a biconvex lens followed by a separate crystalline cone that in some species exhibits a wavy upper profile. A, by now, familiar finding. Even if not made of calcite crystals, the eye of Serolis schythei (Liitken), shown in plate 42, nonetheless gave me a glimpse of what the holochroal eye of a trilobite must have looked like. Its shape and structure beautifully mimic what we had learned about it from the fossil record, but never had seen in its limpid glow. A specimen of Serolis trilobitoides (Eights) is shown in plate 43. Its turret-shaped eyes were not blue anymore, much to my dismay, possibly due to their denaturated state of preservation in ethanol. A radiograph of Serolis schythei is shown in plate 44. What gives away the fact that Serolis is a crustacean and not a trilobite are the double pair of antennae. Trilobites had only one. Most of the ventral appendages, furthermore, are uniramous instead of biramous, among several other differences. Alas, trilobites are indeed extinct. Nevertheless, the overall appearance of the carapace does indeed remind one of, for example, Arctinurus occidentalis Hall (see plate 45), with its falcate pleural spines. 3.3

The

Eyes

of Trilobites

69

PLATE

42

T h e eye o f a specimen o f Liitken,

Serolis schythei

photographed while immersed in

ethanol. T h e eye structure of this living

trilobites. However, the eye lenses are made of organic material in this case, n o t of calcite as the trilobite's. ( x 6 0 )

antarctic isopod ctustacean strikingly

(Specimen loaned by the U.S. National

resembles that of the holochroal eye of

Museum, Washington, D.C.)

70

The Ttilobita

PLATE

43

Dorsal view of a specimen of the isopod

crustacean Serolis

trilobitoides

(Eights),

from the A n t a r c t i c Bransfield Strait, photographed while i m m e r s e d in ethanol. The morphology of this antarctic isopod is reminiscent o f that o f some trilobites, hence the specific n a m e , given by the nineteenth-century naturalist James Eights. (x4.4) (Specimen loaned by the U . S . National M u s e u m , Washington,

3.3

The

Eyes of Trilobites

D.C.)

71

PLATE

44

X-ray radiograph of a

specimen of Serolis

schythei

Liitken,

from

the Straits of Magellan. (x4.5) (Specimen loaned by the U.S. National Museum, Washington, D.C.) Similar life habits and basic metameric body plan may have favored (convergent evolution) the selection of b o d y morphologies similar to those developed independently a n d much earlier by trilobites.

72

The Trilobita

PLATE

45

A perfectly preserved

e x a m p l e of Arctinurus

occidentalis

Hall,

from

the Silurian Rochester S h a l e o f Lockport, New York (xl.49). T h i s is the trilobite that J a m e s Eights alluded to w h e n

trilobitoides

naming

the

isopod of the genus

Serolis,

w h i c h he

discovered on the shores o f the S o u t h Shetland Islands in 1 8 3 0 . (Photographed by the a u t h o r t h r o u g h courtesy of Michael Thomas.)

73

In the meantime, Dr. Clarkson succeeded in interesting the British Antarctic Survey in the task of bringing home several live specimens of Serolis. Observations were made of these gentle creatures frolicking in a tank and staring at the observer with their multifaceted, translucent eyes. A movie was made, one which gave me the eerie feeling of watching a primordial pool crawling with trilobites.

3.4

Enrollment

The great majority of trilobites could roll themselves up so that only the hard carapace was then exposed. In this condition, various types of spines which may have adorned the exoskeleton became functional, protruding from the enrolled body in a defensive manner. There are many examples of this behavior in modern arthropods. A garden variety millipede (Sphaerotherium) is often seen rolled up. In the horseshoe crab, this function is only partially available, since the thorax is fused with the cephalon. The articulation of the abdominal region, then, allows only a flexure, sufficient to draw-up the powerful telson to make a right angle with the rest of the body. Earlier related forms, however, could double up completely. The mechanism of enrollment in trilobites is characteristic of particular phylogenetically related groups, as was recently emphasized by Bergstrom (1973a). The thoracic tergites were clearly engineered to permit this function, and a variety of articulating joints have been identified. To give a most schematic description of the thorax, individual tergites were hinged at two points, often characterized by visible notches, proximal to the axial furrows. Rotation of the tergites around these pivot points would separate the axial rings, exposing the articulating half rings, and at the same time cause the pleural extremities to overlap each other. Special devices, called panderian organs, would, like a doorstop, limit the extent of enrollment. Other devices, the vincular furrows, would ensure a safe interlocking of the pygidial and cephalic margins, as if to discourage casual intruders.

74

The Trilobita

A few major modes of enrollment have been recognized by Bergstrom (1973a). These differ from previous descriptions in that they reflect functional characteristics rather than purely morphological distinctions; these forms of enrollment are sketched in figure 14. Some trilobites were unable to roll up completely, and this mode is called incomplete enrollment. Of those which could roll up completely, a distinction is made between spheroidal and spiral enrollment. Each contains a sequence of subcases. In the spheroidal enrollment, the pygidium comes to rest with its ventral side in contact with the cephalic doublure, not inside it, and the pleurae close the exoskeletal basket laterally. When the pleurae do not wrap around to seal the basket laterally, the enrollment is called cylindrical. When the pygidial termination overlaps the cephalic margin, the mode is called inverted spiral enrollment but is still considered part of the spheroidal main group. The spiral enrollment series contains the case in which the dorsalpzn of the pygidium contacts the ventral side of the cephalon; the uncoiled spiral enrollment, characteristic of the calymenids, where the pygidium is visible even in the enrolled condition; and finally the so-called basket and ///^enrollment. Both spheroidal and spiral enrollment seem to have evolved from the incomplete enrollment of early Cambrian trilobites. Although enrollment most likely represented a defense mechanism, in the long run it may have precipitated the disappearance of the trilobites. It is conceivable that, with the advent of fishes, an enrolled trilobite could have been swallowed more easily than an outstretched one. Fortunately for us, if not for them, enrolled trilobites made better fossils, since the hard carapace is often quite impervious to weathering agents. The Middle Cambrian trilobite Elrathia kingii (Meek) from the Wheeler Formation of Utah offers one of the earliest examples of spheroidal enrollment. In plate 46 we see an enrolled specimen of this species from the cephalic side. The ventral side of the pygidium is seen to protrude beyond the cephalic border, possibly as a result of compression. The entire specimen is, in fact, considerably flattened. The picture shows how the genal spines protrude from the body when the trilobite is enrolled.

FIGURE

14

S c h e m a t i z a t i o n of the types of e n r o l l m e n t in trilobites. (a) I n c o m p l e t e

enrollment,

as in Kjerulfia. (b) C y l i n d r i c a l

enrollment

as in Fallotaspis. (c) Spheroidal

enrollment

proper, as in Asaphus. (d) Inverted spiral e n r o l l m e n t as in

Placoparia.

(e) Spiral e n r o l l m e n t proper,

as in FJlipsocephalus. (f) Uncoiled spiral enrollment, as in

Flexicalymene. (g) Basket and lid enrollment, as in

Harpes.

(Adapted from

Bergstrom,

Another case of spheroidal enrollment, this time in full undistorted relief, is shown in plate 47 (Clarkson 1973a). The trilobite is Encrinurus variolaris (Brongniart), from the Silurian of Dudley, England. Here the closure of the basket is perfect, and one begins to understand why the outline of a trilobite's carapace is often so longitudinally symmetrical. The two halves have to match rather accurately to ensure a complete enclosure in the enrolled condition. In plate 48, parts a, b, c, and d, we see four different

1973a.)

moments in the enrollment process of Phacops rana. In the Devonian Silica Shale at Sylvania, Ohio, this trilobite is commonly found in various postures, from the outstretched to the completely enrolled one. This has nothing to do with incomplete enrollment. The mode describes the ultimate capacity and not the intermediate steps to reach it. Although the character in plate 48 did not quite perform according to the prescription, we are still dealing with a classic example of spheroidal enrollment.

3.4

Enrollment

75

PLATE

46

Enrolled specimen of

Elrathia kingii (Meek), seen from the cephalic side. Middle C a m b r i a n , Wheeler Formation, Utah (xlO). (RLS coll.; now at F M N H . ) This is an early example of spheroidal enrollment.

PLATE

47

A perfect example ot spheroidal e n r o l l m e n t . T h e trilobite is

Encrinurus variolaris (Brongniart), Silurian, Dudley,

England

(x7.6). (Negativeloaned by E. N. K. Clarkson.)

76

The Trilobita

PLATE

48

Various postures in the process of enrollment o f

Phacops rana,

from

the

D e v o n i a n Silica Shale of Sylvania, O h i o (x2.4). (RLS coll.; n o w at F M N H . )

3.4

Enrollment

11

In plates 49 and 50, we encounter the same Dalmanites pratteni Roy that, in section 3.3, was shown to have such special eyes. The specimen is perfectly enrolled and represents an example of spheroidal enrollment in which the pygidium extends quite a way beyond the cephalic border. It must be realized that, due to its construction, the trilobite could not do any better than this in rolling up. The seal was very tight, however, as can be seen in plate 50, in spite of the limitations. An example of uncoiled spiral enrollment is shown in plate 51. The trilobite is Flexicalymene meeki (Foerste), Ordovician, from Waynesville, Ohio. Although the enrollment resembles the spheroidal type, it has been shown by Bergstrom that this particular form is at the end of the evolutionary line of spiral enrollment. The margin of the pygidium is safely interlocked in a groove beneath the frontal cephalic doublure. The lateral cephalic margin contributes to wrap the pleural basket. The expansion of the axial rings and exposure of the articulating half rings is clearly visible in this beautifully preserved trilobite. It is the author's experience that enrollment helped the trilobite survive even as a fossil. The little ball was found intact after having been washed away from its burial sediment and down a steep ravine by a trickle of water. The other trilobites, which were not tightly enrolled (and which were not molts), would irreparably disintegrate given the same exposure.

3.5

Life Habits

An accurate reconstruction of the mode of life of trilobites cannot, of course, be given, since we only know them as fossils. Putting together bits and pieces of evidence, however, as in a detective story, we can arrive at a fairly plausible picture of trilobite habits. First of all, we know that trilobites were exclusively marine animals. Their habitat must have varied over a wide range of conditions, as can be conjectured on the basis of their adaptation. The morphology of the trilobite itself is one of the primary criteria in deducing life habits. Further evidence at our disposal is represented by 78

The Trilobita

faunistic associations, type of bed sediments in which the fossil was located, and ultimately the trilobite's own footprints. The description of the ventral appendage apparatus in Triarthrus •was discussed in section 3.2 in connection with feeding habits. This is an example of the kind of inference that can be made in a particular case. Although this trilobite did not qualify as a predator but as a rather timid smallparticle feeder, there are other trilobites provided with better jaws, and there is evidence that some trilobites actually hunted down their prey in soft-bottom sediments or in the water. In general, trilobites crawled on the sea floor, leaving wellknown trail patterns. In a few cases, the fossil trail leads to the trilobite itself, nested in its own burrow. Because of their many telopodites, actual locomotion for trilobites was a byproduct of sifting and reworking the soft substratum in search for food. Some pulled their bodies sideways in this process; others would use their telopods to burrow more deeply to reach for located prey. In other instances, burrows indicate a good resting or hiding place, or an observation post from which to wait for approaching prey. Burrowing became a mode of life for certain groups of trilobites, which developed a morphology particularly adapted to this purpose, such as a smooth exterior and broad axial lobe. The smooth exterior has the clear implication of reducing friction, and the wide rachis must have housed powerful appendage muscles essential to efficient burrowing. Some illaenid and asaphid trilobites have been found in what is thought to correspond to their life posture. The cephalon would rest on the surface, while the rest of the body projected downwards, as shown in figure 15, following the description by Bergstrom (1973a). Active swimming, on the other hand, must have been the main occupation of other groups of trilobites. In these, the enrollment capacity is reduced, the body is slender and lighter, the pygidium small, the cephalon built hydrodynamically to favor laminar flow. In general, the swimming trilobites had large eyes, with a field of view spanning a circular horizon. P. E. Raymond (1939) must have thought of these characters when he referred to trilobites as the butterflies of the seas. The extreme extension of the visual

PLATE

49

T h e trilobite represented in enrolled position is the same specimen o f

Dalmanites

pratteni

R o y , w h o s e eyes are s h o w n i n plates 3 7 and 38 (x2.3). (Loaned by Field Museum

o f Natural

History.) In this p h o t o g r a p h , the portion o f matrix covering the cephalon and carrying the tip of the pygidium is s h o w n removed from

the

enrolled trilobite. T h e cephalic margin is seen to fit tightly against the ventral side of the pygidium.

3.5 Life Habits

79

PLATE

50

T h e t w o parts o f the specimen in plate 49 are n o w assembled to show h o w the trilobite was originally found. In no way could the trilobite bring the cephalic margin to match the border of the long pygidium. Even so, enrollment provided effective protection, also in view of the presence of a robust telson, n o t preserved here, protruding from the tip of the p y g i d i u m .

80

T h e Trilobita

PLATE

51

A perfectly enrolled specimen o f the O r d o v i c i a n trilobite

Flexicalymene (Foerste)

meeki

from

Waynesville, Ohio ( x 9 ) . ( R L S coll.; n o w at F M N H . )

FIGURE

15

Probable life posture of the illaenid trilobite

Panderia

megalophtalma

Linnarson. (Adapted from Bergstrom 1973a.)

3.5

Life Habits

81

surface to the ventral side, as seen in section 3.3 for Pricycbpyge and others, is telltale evidence of swimming life. If we should learn anything from the horseshoe crab, we might infer that some trilobites liked to swim upside-down. With their carapace functioning as a glider in this attitude, their dorsal eyes could scan the sea floor for prey more efficiently than in the upright position. This, however, is only speculation, after having seen the horseshoe crab, particularly the young, frolic and swim upside-down in a marine tank at Hinds Laboratory for the Geophysical Sciences at the University of Chicago. One can also speculate that the turretlike eyes of many trilobites, protruding as they did from the cephalic surface, could have been very profitably used as watchtowers above the seafloor level and still enable complete concealment of the hunter below a layer of sand. The appearance of a hard carapace itself, the enrollment capabilities, and various forms of spinosity are response to a need for protection against an environment that may have become hostile in many ways. On the whole, we must conclude that 350 million years of survival are good evidence of successful trilobite adaptation.

3.6

Trilobite Classification

The systematization of the fossil record of trilobites into a scheme that reflects the phylogeny and evolution of different stocks in a consistent manner has been the constant preoccupation of paleontologists since the early nineteenth century. Even the most basic subdivision of the material into major groups has been and still is the subject of great controversy. The problem of properly assigning the parenthood and relationships of the fifteen hundred or more genera known today, encompassing some ten thousand different species, is a monumental task indeed. The discovery of particular anatomical or functional features in trilobites has led, in the past, to attempts at classification based on such features. The presence or absence of eyes (Dalman 1827), the eye structure (Emmrich 1839), and the enrollment ability (Milne-Edwards 1840) are only 82

The Trilobita

a few examples of the criteria which played a role in the earliest attempts of classification. The list is long indeed and is discussed in detail in the Treatise (Moore 1959). In the search for a "natural classification," Beecher (1897) believed that the natural sequence of evolutionary events could be unraveled by studying the ontogenetic development of the species (Haeckel's law of ontogenetic recapitulation). This led Beecher to subdivide the trilobites into three main groups on the basis of the pattern of their cephalic sutures. The Proparia and Opisthoparia are two of them, already mentioned in section 3.1; in addition he included a third group, the Hypoparia, which have ventral cephalic sutures. The first two groups have formed the basis of numerous subsequent classifications, though not without objection from other paleontologists. The third group has been rejected altogether. In the more recent attempts at classification, it has been recognized that no individual feature, no matter how significant it may be, is a sufficient guide for classification and that affinities based on collective characteristics must be weighed. The cephalic sutures have always been considered a significant guideline, up to the classification adopted in the Treatise, together with the cephalic axial characters. Further discussion along this line, however, threatens to become too technical and beyond the scope of this work. For this we refer the reader to the Treatise (Moore 1959) and to some of the literature that will be referred to in the following discussion. At the Oslo Trilobite Conference of July 1973, I met Dr. Jan Bergstrom, then at the Department of Historical Geology and Paleontology, University of Lund, Sweden. In a comprehensive survey and innovative study (Bergstrom 1973a), Dr. Bergstrom had just reexamined critically the problem of trilobite classification. In Bergstrom's work, the two main types of trilobite enrollment, spheroidal versus spiral, have been adopted as an additional phylogenetical index that has helped to sort out natural groups when integrated with previously used criteria (Richter 1933; Henningsmoen 1951; Hupe 1953, 1955). As a result, a more consistent picture had emerged—one which consolidated much of the previous proliferation of orders and

suborders, often based on characters of no phylogenetic significance. In many respects, Bergstrom's classification departed radically from the Treatise. Having been brought up to date by Dr. Bergstrom, and still preparing the first edition of this book at that time, I did decide to organize the atlas material according to the newly proposed classification, which recognized nine orders of trilobites. A pictorial illustration of these groupings, adapted from Bergstrom (1973a), was given in figure 14 of my book (Levi-Setti 1975). Modern science is a communal enterprise. In both experimental and theoretical work, novel results and their interpretation must await confirmation and acceptance by a broad community of scholars before becoming established. Unless a consensus is reached, based on uncontroversial evidence, even innovative ideas remain relegated to the realm of speculation. It is now almost twenty years since Bergstrom's revision of trilobite classification first appeared. Alas, a consensus is still wanting. As much as several of its implications have been praised, others have raised objections. For example, recourse to enrollment type as a criterion to sort out one group from another, the Ptychopariida in particular, has not been universally accepted. Morphological characters are still carrying different weight, depending on the school of thought. Dr. Bergstrom has in the meantime updated and refined his original revision and has kindly provided me with a still unpublished sketch representing his current thinking about the phylogeny of trilobite groups and consequent classification at the order and suborder level. This updated classification attempt is schematized in figure 16, where sketches of the trilobites that are representative of each grouping are also shown. It should be noted that the Agnostida are included in this scheme with a question mark: they may not be related to the trilobites in view of their affinities with the Crustacea. This novel scheme has many attractive features and may well form the basis for a classification that may be universally accepted at some future time. Confronted, however, with a still unsettled outcome of many controversies, I had to decide how to reorganize the presentation of the material in the atlas, in a manner

that would survive the test of time and avoid rejection on part of any of the contending and expert factions. Of some help in this dilemma have been the objections to recently advanced trilobite classification schemes that have been raised by my colleague and trilobite scholar Franco Rasetti. In a critical discussion of the problems of trilobite taxonomy (Rasetti 1972), he conceded that, with Jaekel (1909) and Hupe (1953, 1955), only two groups of trilobites, the Miomera (two or three segments) and Polymera (more than three segments) may deserve order ranking. This view inspired me to seek a simple solution. In a brainstorm, I visioned standing in front of an exposure of fossiliferous strata, asking myself what, if any, could be regarded as uncontroversial evidence concerning trilobite phylogeny and evolution. I realized that the revised version of my atlas presentation was layed out for me page by page, layer by layer, forever embedded in the rocks I was staring at. What is uncontroversial in the fossil record is the stratigraphic location of a particular group of life forms in the so-called geological column. Moderately objective is also the recognition of families, encompassing types that are viewed as related at the genus and species level. As a result of the above considerations, I have decided to adopt a chronological sequence in this presentation of the atlas material. The trilobites that populate each geological period will be listed at the family level, without concern about higher taxonomic groupings. Unlike the sequence of strata that we find in sedimentary rocks, however, where the youngest layers are the first to be encountered, I prefer to start at the bottom of the column, showing first the oldest trilobites. The geologic column is not evenly populated. Out of 140 families listed in the Treatise, 92 appear in the Cambrian, while only two remain during the Permian period. A graphic representation of the frequency of occurrence (diversity profile) of trilobite families over the various geologic periods is given in figure 17. It is clear that, even if the catastrophic extinction that is known to have taken place at the Permian-Triassic boundary (Sepkoski 1990) had not occurred, trilobites were well on their way toward a natural disappearance by the end of the Permian. 3.6

Trilobite Classification

83

FIGURE

16

A possible phylogenetic assignment of trilobites to orders and suborders. A r r o w s point to sketches of the typical representatives of the various groups. Interpretation by the author of recenr views expressed by J. Bergstrom (private communication).

84

The Ttilobita

FIGURE

17

Frequency of o c c u r r e n c e (diversity profile)

o f trilobite

families over the geologic c o l u m n .

3.6

Trilobite Classification

85

A Case History:

The Giant Trilobites of Newfoundland

The first edition of my atlas contained a firsthand description of my encounter with the Middle Cambrian Paradoxides beds of Eastern Newfoundland. This field trip was singled out as deserving a special mention in my book, as it represented a significant episode of scientific discovery. It also told an exciting story linking together Carl von Linne's first identified trilobite, Entomolithus paradoxus, the Newfoundland trilobites that I dug out, and spectacular evidence of continental drift. The excitement has not abated in the intervening years. In fact, a systematic study of the strata holding the giant Paradoxides of Newfoundland (Bergstrom and Levi-Setti 1978) has revealed a fascinating sequence of evolutionary events. This time, I felt compelled to present, as a case history, the completion of a story that left a lingering degree of frustrated suspense in Appendix A of my first edition. The story began on a very wet and cold afternoon in the summer of 1974, when I dug out my first giant Paradoxides from the shale beds exposed in the gorge of the Manuels River, on the coast of Conception Bay, in Eastern Newfoundland. The lore of the paleontological discoveries of the last century flashed briefly through my mind. It was uncanny to me that all of a sudden I could hold in my hand such a treasure, only known to me from the sepia lithographs found in the monographs of the last century's masters, describing the trilobites of Bohemia, Sweden, and Wales. I was aware, from the ongoing work at the University of Chicago by A. M. Ziegler and C. R. Scotese (see Scotese and McKerrow 1990, for a recent review), that Maritime North America, including the Avalon Peninsula, South Wales, and Spain were thought to be part of a sequence of land masses facing North Africa in Gondwana (see figure 18), prior to the crunch of Pangea. These would become reassembled as part of Europe and North America in the subsequent drift that created the present Atlantic Ocean. Somehow, a bite of the old world had been carried along by the North 87

American continent, when it split away from Eurasia. The layers I was digging into were deposited during the Middle Cambrian. They are resting on Lower Cambrian sediments and Precambrian rocks, and are in turn buried under Upper Cambrian and Ordovician strata. This entire sequence was transported with minimal disturbance, so that the fossils that these strata contain have retained their original morphology

and are beautifully preserved. The same marine animals that were buried and fossilized along several shorelines of a fragmented proto-Europe can thus be found in rocks of the same age on both sides of the Atlantic. All this I had read about. But now I was holding this fantastic trilobite, bright yellow and red, even more so under the rain, that was once buried in a very distant and very old Europe. I could touch the amazing reality of continental drift.

FIGURE

18

Computer-generated

reconstruction

of

the distribution of land masses in the M i d d l e C a m b r i a n . O b l i q u e polar view, showing the c o m p o n e n t parts of G o n d w a n a and Baltica that became M a r i t i m e N o r t h America and Europe after the assembly and subsequent breakup of the

supercontinent

Pangea. These components

are

s h o w n with presentday coastline contours, to facilitate recognition.

The

actual contours of the Middle

Cambrian

coastlines may have differed in details from the present configuration.

This

illustration was adapted from a custom-generated m a p kindly provided by C h r i s t o p h e r R. Scotese, Paleomap Project, University of Texas at A r l i n g t o n .

88

The Trilobita

In the following days, determined to explore further the content of the strata that yielded my first Paradoxides, I started quarrying the shale beds layer by layer. In this first attempt at a stratigraphic survey, I could ascertain that I had, in my first encounter, hit upon by accident on a narrow band of layers that contained an extraordinary abundance of complete exuviae. These giant trilobites, up to one foot long and brightly colored in yellow, orange, and red (the colors of iron oxides coating the carapace), were too large for the average size slab to contain. It is a well known theorem of fossil collecting that the most interesting finds seem always to occur near the edges of the slab being uncovered, and continue beyond. With trilobites this large, this frustration became the norm, and it was a scramble to extract and split the adjoining shale slabs, in the hope of finding the other half of a trilobite, and restore the broken specimen to its former integrity (thanks to the fissility of the shale containing large exuviae, this approach was surprisingly successful). I did note then that my complete Paradoxides had a broad pygidium, expanding distally in a spatulalike, trapezoidal shape (see plate A l ) . At first I did not make much of it, but then I noticed that other specimens, collected above my favorite layer, were somewhat different, with a pygidium that tapered distally instead of broadening (see plate A2). Back home, I poured over the literature to find out what was known about the local fauna. The work of many paleontologists, reviewed most recently by Hutchinson (1962), seemed to indicate a close correspondence of the three faunal zones of the Middle Cambrian of the Atlantic province (in ascending order, the Paradoxides bennetti, Paradoxides hicksi, and Paradoxides davidis zones), found at Manuels, with those known from, for example, Wales (Howell 1925). In Scandinavia, a further faunal zone, above the P. davidis, (alias the Ptychagnostus punctuosus Zone) is recognized: that of Paradoxides forchhammeri. In Eastern Newfoundland, however, the trilobites associated with the P. forchhammeri zone have been found (Hutchinson 1962), but not Paradoxides forchhammeri itself. I then compared my findings with the reconstructions and photographs of the Paradoxides of England and

Scandinavia (Angelin 1851-78; Brogger 1878; Lake 1935). Definitely, I concluded, I had been digging in the Paradoxides davidis levels, and many of my trilobites matched perfectly with the description given by Salter in 1863 of the trilobites he had discovered on the impervious cliffs at St. David's, in Wales (see plate A3). In particular, the pygidium of P. davidis does taper and is quite slender. However, the abundant complete exuviae I had found concentrated in a narrow band, all exhibiting the trapezoidally shaped pygidium, seemed at first to match reasonably well the description of Paradoxides forchhammeri Angelin, the trilobite that was missing at Manuels. Although I did realize that if indeed I had stumbled upon this trilobite, it did occur in the wrong place (embedded within the P. davidis zone and not above it), I naively resisted the notion that I might have discovered something new. Trying to fit my findings with known species, and ignoring what stratigraphy was telling me, I presented my giant trilobites in the first edition of the atlas, as possible examples of P. forchhammeri. After all, it is not uncommon for deposited sediments to be turned upside down by tectonic events. At Manuels, however, all evidence points to an undisturbed series of sediments. As it turned out, my initial assumption was in error, I had indeed found something else, something new. The situation was clarified after another field trip to Manuels, this time accompanied by my friend Jan Bergstrom, who was visiting at Memorial University of Newfoundland. Beating the rainy season, we surveyed the Manuels beds together, this time with a yardstick. We located the top of the P. davidis zone, characterized by an ubiquitous layer of phosphatic limestone, and went down layer by layer, recording the depth and nature of all the trilobites we found. After a depth of eight feet, populated primarily by the classical variety of P. davidis, my anomalous beds appeared again and yielded more of the wide-tailed complete Paradoxides (see plates A4, A5), at the exclusion of the previous kind. This time we ascertained that the layer containing this trilobite was only about sixteen inches thick and that the classical variety reappeared again beneath it, after some thickness of barren strata (see plate A6). Another important finding The

Giant

Trilobites

of Newfoundbind

89

PLATE

Al

PARADOXIDES Brongniart,

Paradoxides trapezopyge

1822.

davidis Bergstrom

and Levi-Setti, Middle C a m b r i a n , east side of Manuels River, Manuels, Newfoundland ( x l . 8 ) . This specimen represents the first encounter by the author with the

Paradoxides

fauna a t

Manuels. It was tentatively identified, in the first edition of this atlas, (Levi-Setti 1 9 7 5 , pi. A 4 ) as

Paradoxides forchbammeri

Angelin.

T h e trapezoidal spatulated pygidium and the diverging last pair of pleurae are a distinctive feature of this trilobite, latet described by Bergstrom and Levi-Setti ( 1 9 7 8 ) as a new subspecies of

P. davidis S a l t e r Print from

Kodachrome

slide. (RLS coll., id.; now at F M N H . )

90

T h e Trilobita

PLATE

A2

Paradoxides

davidis

davidis Salter,

Middle

C a m b r i a n , west side of M a n u e l s River, Manuels, Newfoundland

(x0.65).

T h i s large trilobite, originally discovered at St. David's, W a l e s (see plate A 3 ) , represents the ancestral stock (phenotype) o f this genus found in Newfoundland.

Note

the tapered p y g i d i u m . This specimen is thought to represent the vestiges of a living trilobite, in contrast to the more c o m m o n l y found exuviae. (RLS coll., id.)

The

Giant

Trilobites

of Newfoundland

91

PLATE

A3

Cast of the type specimen of

Paradoxides

davidis davidis

Salter, M i d d l e Cambrian, from Porth-y-rhaw, St. Davids, Pembrokeshire, W a l e s ( x 0 . 5 2 ) , reproduced from Bergstrom and Levi-Setti 1 9 7 8 , pi. 2, fig. 1 (by permission of Geologica et Palaeontologica, Lahnberge, Marburg/Lahn). This specimen is distorted by tectonic shear, h o w e v e r , the similarities with the Newfoundland specimen o f the preceding plate can be ascertained. (Original at the Natural History M u s e u m , L o n d o n , cast k i n d l y supplied by Richard A. Fortey.)

92

The Ttilobita

PLATE

A4

A c o m p l e t e specimen of

Paradoxides

davidis

trapezopyge

Bergstrom a n d Levi-Setti, M i d d l e C a m b r i a n , east side o f M a n u e l s River, M a n u e l s , Newfoundland (x0.94). The external impression of a h y p o s t o m e , possibly belonging to the same individual, is superposed to the posterior left side of the thorax. M i n o r reconstruction of the left librigena was p e r f o r m e d w i t h the aid of cast from external impression. T h i s m u t a n t form

of P. davidis is f o u n d in a narrow band of strata, preceded and followed by strata bearing the ancestral phenotype. Specimen whitened with magnesium oxide. (RLS coll., id.)

The

Giant

Trilobites

of Newfoundland

93

PLATE

A5

A n o t h e r example of complete exuviae of

Paradoxides trapezopyge

davidis Bergstrom

and Levi-Setti, extracted in the same vicinity as the specimen in plate A4 ( x l . l ) . It was formerly tentatively identified as

P. forchhammeri

(Levi-

Setti 1 9 7 5 , pi. A 7 ) . Both free cheeks are displaced from their original setting; o n e is overturned. T h e cranidium is partly crushed, exposing the underlying h y p o s t o m a , also diplaced. T h i s specimen was the prize finding of Emile LeviSetti at age 9, in September 1 9 7 4 , w h e n he was given the chance to q u a r r y a slab of his o w n . (RLS coll., id.)

94

The Trilobita

PLATE

A6

Paradoxides davidis davidis Salter, west

side

o f M a n u e l s Rivet, Middle Cambrian, Manuels, Newfoundland (x2.3). T h i s j u v e n i l e trilobite, representing a late meraspid stage, originates from the l o w e r m o s t levels of the

P. davidis

beds at this

locality. It exhibits seventeen thoracic segments, versus the n o r m of t w e n t y for adult individuals. T h e latter t w o segments of the thorax, and the attached pygidium, are displaced forward, overlapping the thirteenth to fifteenth segment. T h e glabella is crushed, exposing the underlying hypostoma. T h e surface of the entire exuviae is still partially covered by a bright yellow patina of limonite. (RLS coll., id.)

The

Giant

Trilobites

of Newfoundland

95

emerged: a different variant to P. davidis appeared in a narrow layer located at about the six-foot level. This trilobite has very short pleural spines and a coarsely granulose carapace (see plate A7), while all other characters matched those of P. davidis. A picture of successive faunal replacements of P. davidis by mutant forms, intercalated with the reappearance of the ancestral form, began to emerge. To put to rest the previous identification of my wide-tailed beasts with P. forchhammeri, we obtained samples of cranidia and pygidia of this trilobite from collections of Scandinavian type specimens. While the pygidia matched quite closely, the cranidia of the Newfoundland find did lack, as is the case for P. davidis, two cranidial furrows that were present in the true P. forchhammeri. The stratigraphy did not fail. It was time to name our new trilobites as subspecies of P. davidis. The classic type had to be called P. davidis davidis. We did compare the Newfoundland specimens with casts of the British types and we concluded to our satisfaction that indeed they were examples of the same trilobite. We then named the short-spined mutant Paradoxides davidis brevispinus, the wide-tailed mutant Paradoxides davidis trapezopyge, and finally a hybrid occurring at the interface between the P. d. trapezopyge and the P. d. davidis levels as P. d. intermedius (see plate A8). The reconstructions of the three principal subspecies of P. davidis, and that of P. forchhammeri, are shown in figure 19.1 should mention that these assignments were based not simply on the appearance of a selected type specimen, as is often done, but on the basis of a statistical morphometric study of various body parts involving several hundred specimens. The characters of an entire population must be known to assess the range of variability of particular morphological indices (a physicist's input here) and avoid the pitfall of naming a new species or subspecies on the basis of individuals exhibiting characters at the tails of a normal population, but still compatible with it. This meant much collecting, preparing, and measuring trilobite's body parts. Another investigation that I undertook was the determination, by X-ray powder diffraction, of the clay minerals making up the shale at various depths. This could provide some clue regarding 96

The Trilobita

changes in the environment that could be related with the abrupt faunal changes that we observed. Indeed, the inferred changes in the depositional environment did correspond to the appearance of our mutants, and also with the repeated reappearance of P. davidis davidis. What was left to be done was to make sense of what we had discovered. It did involve evolution and speciation, in a rather unconventional time sequence, however, since we did find the ancestor at both ends. Can evolution proceed backwards as well as forward? As we discussed at length in our paper (Bergstrom and LeviSetd 1978), our observations find a natural and convincing interpretation in terms of the evolutionary model of punctuated equilibria, advanced by Eldredge and Gould (1972) and further illustrated by the same authors (Gould and Eldredge 1977). Quoting from these auhors, "The norm for a species during the heyday of its existence as a large population is morphological stasis, minor non-directional fluctuations in form, or minor directional change bearing no relationship to pathways of alteration in subsequent daughter species. In the local stratigraphic section we expect no slow and steady transition, but a break with essentially sudden replacement of ancestor by descendant: this break may record the extinction or migration of a parental species and the immigration of a successful descendant rapidly evolved elsewhere in a small peripheral isolated population." Not only does this prediction of how speciation may appear in the fossil record fit what we have seen at Manuels, we may now add a plausible corollary to the above model: that the replacement of descendant by the ancestor may also occur. In other words, we now have evidence that the mutant species is not always successful and that a local return of the ancestor may wipe out the offspring. Evolution can take steps backwards, sometimes. Let us examine the scenario that would give rise to our Paradoxides evolutionary drama. Newfoundland at that time was not far off the coast of Spain, along an arc extending to Wales and Maritime North America, as shown in figure 18. The main gene pool of Paradoxides davidis davidis must have populated this extended coastline and, at a deposition time corresponding to the lowest levels explored, it was

PLATE

A7

A n o t h e r m u t a n t form o f

P. davidis

at

Manuels is represented by P. davidis

brevispinus

Bergstrom

and

Levi-Setti.

pleural spines and enhanced tubercular

Lahn.) O n c e again, this m u t a n t appears

o r n a m e n t a t i o n ( x l . 7 ) . (From Bergstrom

confined to a n a r r o w bed w i t h i n the

P. davidis davidis.

and Levi-Setti 1 9 7 8 , pi. 9, fig. 5, by

layers p o p u l a t e d b y

Exuviae of this new subspecies are s h o w n

permission of Geologica et

(RLS coll.; n o w at F M N H . )

in this plate, characterized by very short

Palaeontologica, Lahnberge, Marburg/

The Giant

Trilobites of Newfoundland

97

FIGURE

19

Reconstruction of the principal subspecies

conjectural, hence the interruption in

of Paradoxides davidis: (a) P. d. davidis, (b) P. d. trapezopyge, (c) P. d. brevispinus.

the assumed sequence of t w e n t y tergites.

Also s h o w n is a reconstruction of

by permission of

Paradoxides forchhammeri (d). The outline

Palaeontologica, Lahnberge, Marburg/

of the thoracic region of the latter is

Lahn.)

98

The Trilobita

(From Bergstrom a n d Levi-Setti 1 9 7 8 , Geologica et

The

Giant

Trilobites of Newfoundland

99

PLATE

A8

A third mutant form of P.

davidis is found at the interface between the bed

containing P. davidis davidis and the underlying bed with

P. davidis trapezopyge. It was named P. davidis intermedins, to stress the hybrid character of the pygidium, intermediate between those of the t w o bordering subspecies. This plate shows nearly c o m p l e t e exuviae of this new subspecies ( x l . 2 ) . (From Bergstrom and Levi-Setti 1 9 7 8 , pi. 6 , f i g . 6 , by permission of G e o l o g i c a et Palaeontologica, Lahnberge, Marburg/Lahn). ( R L S coll.; now at F M N H . )

1 00

The Trilobita

certainly present at Manuels. At some time, possibly due to a sudden bathymetric shift, the local population was cut off from the main pool. Mutation and natural selection in the Newfoundland isolate led to the complete dominance of the subspecies P. davidis trapezopyge. Subsequently, due to a reverse bathymetric shift, connection with the main gene pool was reestablished. What we see then is a brief period of interbreeding (P. d. intermedins) until the local mutant was replaced completely by the ancestral ilk. Then a period of stasis intervened, followed by a new episode of isolation, giving rise to P. d. brevispinus. But shortly afterwards (we are always talking millions of years), the ancestor reappears and wipes out once again the unfortunate mutant. The end eventually occurs also for the dominant ancestor, at the top of the P. davidis zone, coincidentally with evidence of widespread volcanic activity. Remarkably, several of the characters developed by the two mutant subspecies, namely the trapezoidal pygidium and the granulose carapace surface, will reappear at a later time, incorporated in a veritable new species, Paradoxides forchhammeri. Were the mutants of P. davidis less fit for survival than the ancestor? Possibly so: an unusually high rate of occurrence of genetic malformations (teratologies) was observed for the mutant subspecies. An example is shown in plate A9.

This concludes the case history, except for the Linnaean connection. In the first edition of the atlas, I reproduced the reconstruction of Carl von Linne's Enthomolithus paradoxus, as given by Angelin (1878). I did remark at that time that a remarkable resemblance did exist between this and our new trilobite Paradoxides davidis trapezopyge. In the following years, Jan Bergstrom discovered that an actual photograph of the original Linnaean specimen had been published by Nathorst (1907). Pursuing the quest for a fresh look at this historical trilobite, Jan Bergstrom finally located the specimen at the Mineralogical Museum of the University of Copenhagen, took pictures, and kindly provided me with the photograph reproduced in plate A10. This trilobite has twenty tergites, as P. davidis does, and indeed a trapezoidal pygidium. Its glabella is crushed, showing the underlying labral plate, as in many of our specimens. The cranidium and free cheeks are displaced toward the thorax, overlapping the first two tergites. All visible characters match quite well those of my first giant Paradoxides, found on my first trip to Manuels on a wet afternoon. Was it P. davidis trapezopyge Bergstrom and Levi-Setti 1978 that I stumbled upon, or was it Enthomolithus paradoxus Linnaeus 1753? This we may never establish beyond doubt. Nevertheless I am left with the feeling of having shared with Linnaeus the experience of our first encounter with the same giant trilobite.

The

Giant

Trilobites

of Newfoundland

PLATE Several

A9 malformations

(teratologies) affect the exuviae of this example of P.

davidis trapezopyge

Bergstrom

and Levi-Setti ( x l . 8 ) . (From Bergstrom and Levi-Setti 1 9 7 8 , pi. 9, figs. 1 - 3 , by permission of Geologica et Palaeontologica, Lahnberge, Marburg/Lahn.) T h e deformities involve the fusion of tergites 1 and 2, an anomalous n u m b e r of tergites ( 1 8 instead of 2 0 ) , a gross malformation of the fight pleura of tergites 7 t h r o u g h 1 1 , and atrophic left 1 5 t h and right 1 8 t h pleural spines. (RLS coll., id.; n o w at FMNH.)

102

The Trilobita

PLATE

A 1 0

Entomolithus

paradoxus

Linnaeus ( x 0 . 4 6 ) . This is the trilobite specimen described by C a r l v o n L i n n e (Linnaeus 1 7 5 9 ) , n o w regarded a s a n

example of Paradoxides paradoxissimus (Wahlenberg). T h e trapezoidal p y g i d i u m and the strongly diverging last pair of pleurae a n d pleural spines suggest, h o w e v e r , closer

affinity

to

P. forchhammeri

(among the Swedish

Paradoxides

only). In this

example, the entire c r a n i d i u m is displaced to cover the first few thoracic segments. A photograph of this specimen was originally published by Nathorst ( 1 9 0 7 ) . This is a reproduction from a m o d e r n Polaroid picture of the original taken by Jan Bergstrom at the Mineralogical M u s e u m of the University of C o p e n h a g e n , Denmark.

The

Giant

Trilobites

of Newfoundland

103

Photographic Techniques and Specimen Preparation

Very little can be added to the description of the physical principles involved and the technique of fossil photography given by Franco Rasetti in the Handbook of Paleontological Techniques (Kummel and Raup 1965). As a physicist, I have appreciated Rasetti's suggestions tremendously and have endeavored to translate them into practice as much as possible. Although it is not generally recommended for high resolution work, 35mm. film was used for all of the photographs. The use of slow, fine-grained film and careful development procedure has, however, enabled enlargements to 8" X 10" print size without loss of quality. For this purpose, a now irreplaceable negative film was used—ADOX KB 14, 20 ASA. Many of the photographs that appeared in the first edition of this atlas were taken before this film was discontinued. For most of the new plates presented here, the film of choice was Kodak Technical Pan, 25 ASA. The ADOX film was developed in AGFA Rodinal developer, diluted 1:100, for 20 minutes at 20° C. The Kodak film was developed with the same developer, dilution, and temperature, but for only 5 minutes. Microscopic examination of the developed image in these films showed a grain size and lack of graininess (coalescence of groups grains) much superior to any other presently available negative material. A recently introduced color negative film, Kodak Ektar 25 ASA, was occasionally used for black and white printing with excellent results. This, however, required use of high contrast paper. For those specimens photographed on location, often in less than optimal lighting conditions, the Ilford films XP1, XP2, and Delta, 400 ASA, were used with some success. I seem to have found my best solutions to the problems of fossil photography in the use of materials and equipment now discontinued. Among the latter is an item responsible for the best focusing in the pictures presented here, a Leitz 105

device called Reprovit. It consists of a sliding mount, carrying the camera body (a Leica) and a ground glass placed at the level of the film plane in the camera. Beneath the sliding mount, the lens is usually carried by a focusing extension or, as modified by the author, a bellows controlled by rack and pinion motion. Focusing is achieved by projecting the image of a flat marker, encased in the ground glass, onto the surface of a specimen, as if the device were a photographic enlarger. By doing so, the field of view seen by the lens is also illuminated, and the specimen can be properly centered or positioned. The lens stop is then closed to the desired setting, the projector removed and the camera made to slide into position above the lens. The focal plane shutter is open and the exposure takes place by switching the illuminating lamps on and ofT. Depending on the size of the specimen and depth of field required, different lenses were used, usually stopped to the largest f: number. Among the lenses used are Leitz Elmar 50mm (f:16), Schneider Componon (f:22), and Schneider Symmar 135 mm (f:45). The latter two are projection lenses, originating from the author's bubble chamber film scanning projectors. For printing, a Leitz enlarger with 50mm Elmar lens was used, as well as an Omega enlarger with a Schneider 80mm lens. The enlargements for the first edition were printed on Agfa Brovira paper, mostly extra-hard, developed in Kodak Dektol developer diluted 1:2. All new pictures presented here were printed on Ilford Multigrade paper. Specimen illumination was achieved generally with two high-intensity desk miniature lamps, at times with a parallel beam of light from a microscope illuminator, and occasionally with a ring fluorescent lamp. The paleontological convention of illuminating samples from the Northwest has not been adhered to in several cases, when lighting from other directions may have led to better photography. As described in Rasetti's article, the immersion of specimens in xylene often enhances the contrast between the specimen and matrix. This technique was employed to particular advantage when mineralized or pigmented details were embedded in surface incrustations of calcite or quartz. The optical contact between these minerals and the xylene, 106

The Trilobita

having approximately the same refractive index, enables light to penetrate unreflected and unrefracted through the surface layers, revealing the underlying structure of the fossil. Immersion in xylene yielded quite spectacular results also in enhancing the contrast of the pyritized appendages of the Utica Shale trilobites. In several instances, when color contrast existed between specimen and matrix, it was found worthwhile to take a color photograph on a Kodachrome slide and then print directly from the slide. The resulting print is a negative, but this is irrelevant in fossil photography. The high contrast factor of Kodachrome (not Ektachrome) yields extremely sharp prints. Valuable sources of further technical information on the photography of small objects in general are two Kodak technical publication manuals—N 12A on close-up photography, and N 12B on photomacrography.

Specimen Preparation The preparation of the specimens for photography has often been a painstaking operation, usually carried out under a stereoscopic microscope. Some of the methods adopted followed the advice of Rasetti and Palmer in the Handbook of Paleontological Techniques (Kummel and Raup 1965). Different procedures have been employed in exposing trilobites preserved in shale from those preserved in, say, limestone. Soft shale—for example, the Silica Shale—disintegrates in water, and so does the trilobite. Here the matrix is easily removed to expose the trilobite details by using a variety of scraping tools and soft bristle brushes. Metal brushes are to be avoided at all times. The use of varnish or shellac to waterproof or consolidate the exoskeleton has been avoided in order not to cause unwanted reflections. Repairs of flaked parts have been successfully obtained by using Duco cement diluted in amyl acetate. (The suggestion is from Rasetti and Palmer.) This diluted glue dries without leaving a glossy surface. Final cleaning of the trilobite was obtained by using a moist cloth. A somewhat harder shale is the Waldron Shale or Wheeler Shale. Here prolonged soaking in a solution of

Quaternary O (Geigy Industrial Chemicals) is the best preparation. Caution should be taken to avoid any procedure which could scratch the calcite of the carapace. Unless the specimen is whitened, chisel marks are very disturbing to good photography. In general, care also has been taken to present the matrix in a fairly natural condition. The hardest preparation is that of trilobites in limestone. Only rarely, as in specimens from Grafton, Illinois, does the separation of trilobite and matrix obtain naturally. More often the trilobite can only be exposed by resorting to rather arduous chiseling and grinding and to the aid of vibro-tools. In these cases, the matrix surface is usually left scarred by white marks. In such instances, careful illumination has been used in order to form shadows on the unwanted scratches.

The trilobites presented are, in many instances, the only survivors of unsuccessful attempts to prepare adequate samples from a much larger initial body of specimens. For obvious reasons, the specimens borrowed from museums or from private collections had to be left untouched. It is standard procedure in paleontology to whiten fossils for photography. This approach has been followed here occasionally by exposing the specimen to magnesium oxide vapors from a burning magnesium ribbon (Rasetti 1947). In conclusion, this book contains a spectrum of both orthodox and unorthodox approaches to the photography of fossils. These notes contain no implication whatsoever that any of the methods employed here are preferable to the accepted standards in paleontology. My personal tastes and the means at my disposal have been the only guidelines here.

Photographic

Techniques

and

Specimen

Preparation

107

ATLAS OF

TRILOBITE

PHOTOGRAPHS

The atlas is organized in six main sections corresponding to the geologic periods that contain the fossil record of trilobites, from their appearance at the beginning of the Cambrian to their demise at the end of the Permian. Each section is preceded by a listing of the trilobite families that are found in that particular geologic time interval, in order of appearance, following the information available in the Treatise (Moore 1959). The genuses illustrated by atlas plates are indicated in parentheses for each family. Since a time sequence translates into a depth series in the geologic column, the customary subdivisions of each geologic period into Early, Middle, and Late epochs correspond to Lower, Middle, and Upper series when referring to the stratigraphic position. Thus the latter nomenclature will be used in indicating the survival time span for each family. A number of families survived over more than one particular geologic period. Arrows pointing to the left or right of each series name indicate earlier or later presence of that particular family in the geologic column. Unfortunately, it is not possible to maintain a chronological ordering of the first appearance of each family and at the same time preserve evolutionary continuity when a particular family is present throughout more than one period. In a few of these cases, then, related forms will appear at different locations in the atlas. The captions accompanying the plates contain detailed taxonomic and stratigraphic information but also stress general and often nontechnical features. It should be emphasized that the vastness and diversity of the fossil record of trilobites prevents an exhaustive coverage of the material, with large format photographs. The visual impact of this choice of presentation is the trade-off of a more encyclopedic approach, beyond the scope of my atlas. The main preoccupation here has been that of presenting more vivid, real life trilobite reproductions, than, for example, those provided by the accurate drawings which adorn 109

the professional literature. Furthermore, trilobites are generally small, and reproductions in natural size tend to become insignificant. Adequate enlargement often reveals a lot more than the unaided eye can see. In selecting the photographs to be included in each section, I have been partial to their aesthetic appeal, and I have also taken irreverent liberties in choosing mostly perfect specimens for presentation and in deliberately using nonconventional photographic techniques. Since, with rare exceptions, I have preferred to rely on my own photography, the trilobite specimens had to be available in my laboratory. Thus a large number of specimens originate from my own personal collection or were borrowed from various museums. On a few occasions, I transported my photographic equipment on location to have direct access to exceptional private collections. Although, for this revised edition, I made an effort to limit repetition and broaden the range of coverage, the atlas is still based on a very limited selection of trilobite specimens that I was able to photograph on my spare time. One of my wishes has been to point out, with these pictures, that dinosaurs were not the only prehistoric animals that inspire awe and fascination. Trilobites tell us of an earlier world, perhaps less threatening, when life on earth could still explode into a myriad of new, unseen, uncounted forms discovering their own way to survive. They tell us of a time scale that transcends our power of imagination and how these immense spans of time allowed them to experiment and develop models and methods of biological function, vision perhaps above all, that are as awe-inspiring as the advent of giant behemoths.

Atlas of Trilobite Photographs

Explanatory Note on Abbreviations and Nomenclature Adopted in the Adas The following abbreviation code has been adopted in the atlas to indicate geologic column subdivision and other abbreviations used in the captions to atlas plates. This tabulation also includes the key to the word terminations used to denote taxonomic classification in invertebrate paleontology. In each legend the generic attribution (genus) is indicated first, followed by synonyms, if any, or other specifications. The combination of generic and trivial name for the species represented in the plate is indicated in italics. Whenever more than one species belonging to the same genus are presented in sequence, only the legend of the first plate of the sequence carries the generic name, author, and date. With the exceptions indicated in the captions, all photographs have been prepared by the author. Photographic techniques and specimen preparation are described in Appendix B of chapter 3. Unless a special technique is mentioned, the specimens have been photographed in air and uncoated. Many of the specimens originate from the University of Chicago Walker Museum, now transferred to the Field Museum of Natural History in Chicago. The disposition of this material and that from other museums will be indicated in abbreviated form (see code). The names of private collectors who have either contributed photographs, loaned specimens to the author, or permitted the author to photograph specimens on location will be acknowledged in the captions. The responsibility for the classification and identification of specimens indicated by the term RLS id. rests solely with the author.

ABBREVIATION

Geologic

CODE

Column

System CAMBRIAN

CAM.

ORDOVICIAN

ORD.

SILURIAN

SIL.

DEVONIAN

DEV.

CARBONIFEROUS

CARB.

Mississippian

Miss.

Pennsylvanian

Perm.

PERMIAN

PERM.

Series Lower, Middle, Upper

L., M., U.

Abbreviations

and

Nomenclature

Adopted

in

the

Atlas

111

Miscellaneous Identified

id.

Collection

coll.

Field Museum of Natural History, Chicago

FMNH

Geological Survey of Sweden, Uppsala, Sweden

GSS

Grant Institute of Geology, University of Edinburgh

GRI

Museum of Comparative Zoology, Harvard University

MCZ

Peabody Museum, Yale University

YPM

Swedish Museum of Natural History, Stockholm, Sweden

SMNH

University of Chicago, Walker Museum

UCWM

U.S. National Museum, Washington, D.C.

USNM

Taxonomic Nomenclature

Taxa

Termination

Example

Class

a

Trilobita

Order

ida

Phacopida

Suborder

ina

Phacopina

Superfamily

acea

Phacopacea

Family

idae

Phacopidae

Subfamily

inae

Phacopinae

Genus Species Subspecies

Phacops rana milleri

CLASS

TRILOBITA

WALCH,

4.1

1771

Cambrian Families

OLENELLIDAE

Vogdes, 1893

L.Cam. (Olenellus,

Bristolia) DAGUINASPIDIDAE

CALLAVIIDAE

Hupe, 1953

Daguinaspidinae Hupe, 1953

L.Cam.

Fallotaspidinae Hupe, 1953

L.Cam.

Poulsen, 1959 Callaviinae Poulsen, 1959

L.Cam.

Archaeaspidinae Repina, 1979

L.Cam.

HOLMIIDAE

Hupe, 1953

L.Cam. (Andalusiana)

WANNERIIDAE

Hupe, 1953

L.Cam. (Wanneria)

NEVADIIDAE

Hupe, 1953 Nevadiinae Hupe, 1953

L.Cam.

Judomiinae Repina, 1979

L.Cam.

(?) Neltneriinae Hupe, 1953

L.Cam.

Cambrian

Families

REDLICHIIDAE

Poulsen, 1927 Redlichiinae Poulsen, 1927

L.Cam. (Redlichia)

Pararedlichiinae Hupe, 1953

L.Cam.

Neoredlichiinae Hupe, 1953

L.Cam.

Wudngaspidinae Chang, 1966

L.Cam.

Hupe, 1953

L.Cam.

GIGANTOPYGIDAE

Harrington, 1959

L.Cam.

DESPUJOLSIIDAE

Harrington, 1959

SAUKIANDIDAE

Despujolsiinae Harrington, 1959

L.Cam.

Resseropinae Chang, 1966

L.Cam.

PROTOLENIDAE Richter and Richter, 1948 Protoleninae Richter and Richter, 1948

L.Cam.

Myopsoleninae Hupe, 1953

L.Cam.

Bigotininae Hupe, 1953

L.Cam.

Termierellinae Hupe, 1953

L.Cam.

Palaeoleninae Hupe, 1953

L.Cam.

Lermontoviinae Suvorova, 1956

L.Cam.

Bergeroniellinae Repina, 1966

L.Cam.

ALDONAIIDAE

Hupe, 1953

L.Cam.

YUNNANOCEPHALIDAE

Hupe, 1953

L.Cam.

Whitehouse, 1939

L.Cam.

Suvorova, 1958

L.Cam.

METADOXIDIDAE JAKUTIDAE DOLEROLENIDAE

114

Atlas o f Trilobite Photographs

Kobayashi, 1951

L.Cam. (Dolerolenus)

HICKSIIDAE

Hupe, 1953

L.Cam.

ABADIELLIDAE

Hupe, 1953

L.Cam.--M.Cam.

ELLIPSOCEPHALIDAE

Matthew, 1887

Ellipsocephalinae Matthew, 1887

L. Cam.—M. Cam. (Ellipsocephalus)

Strenuellinae Hupe, 1953

L.Cam.-M.Cam.

Kingaspidinae Hupe, 1953

L.Cam.

Antatlasiinae Hupe, 1953

L.Cam.

ORYCTOCEPHALIDAE

Beecher, 1897

Lancastriinae Kobayashi, 1935

L.Cam.

Cheiruroidinae Kobayashi, 1935

L.Cam.

Oryctocephalinae Beecher, 1897

M.Cam.

Oryctocarinae Hupe, 1955

M.Cam.

Tonkinellinae Reed, 1935

M.Cam.

CONOCORYPHIDAE

Angelin, 1854

L.Cam—M.Cam. (Conocoryphe, Bailiella, Ctenocephalus)

Swinnerton, 1915

L.Cam.-M.Cam. (Zachantoides)

ANDRARINIDAE

Raymond, 1937

L.Cam.-M.Cam.

DORYPYGIDAE

Kobayashi, 1935

L. Cam .-M. Cam. (Olenoides)-\J. Cam.

ZACANTHOIDIDAE

DOLICHOMETOPIDAE

OGYGOPSIDAE CORYNEXOCHIDAE BATHYNOTIDAE DINESIDAE

Walcott, 1916

Rasetti, 1951

L.Cam.-M.Cam. (Bathyuriscus, Hemyrhodon)-\i. Cam. M.Cam. (Ogygopsis)

Angelin, 1854

M.Cam.

Hupe, 1953

M.Cam.

Lermontova, 1940

M.Cam.

Cambrian

Families

PARADOXIDIDAE

Hawle and Corda, 1847

Paradoxidinae Hawle and Corda, 1847

M.Cam. (Paradoxides, Eccaparadoxides, Acadoparadoxides)

Xystridurinae Whitehouse, 1847

M.Cam. (Xystridura)

Centropleurinae Angelin, 1854

M.Cam. (Anopolenus)

YINITIDAE

Hupe, 1953

(= DREPANOPYGIDAE Lu, 1961) Yinitinae Hupe, 1953

M.Cam.

Drepanopyginae Lu, 1961

M.Cam.

Whitehouse, 1939

M.Cam.

NEPEIDAE AGRAULIDAE

Raymond, 1913

BOLASPIDIDAE

Howell, 1959

ALOKISTOCARIDAE

PTYCHOPARIIDAE

Resser, 1939

Matthew, 1887

Ptychopariinae Matthew, 1887 Periommellinae Rasetti, 1955 Antagminae Hupe, 1953

M.Cam. (Agraulos) M.Cam. L.Cam.—M.Cam. (Alokistocare, Piochaspis, Elrathia)-U .Cam. L.Cam. L.Cam.-M. Cam. (Ptychoparia)-\J. Cam. L.Cam. L.Cam.-M.Cam.

Hupe, 1953

M.Cam.-U.Cam.

EMMRICHELLIDAE

Kobayashi, 1935

M.Cam.-U.Cam.

LIOSTRACINIDAE

Raymond, 1937

Conokephalininae

Liostracininae Raymond, 1937

M.Cam.-U.Cam.

Doremataspidinae Opik, 1967

M. Cam .-U. Cam.

CONOKEPHALINIDAE CREPICEPHALIDAE PAPYRJASPIDIDAE

Hupe, 1953

M.Cam.

Kobayashi, 1935

M.Cam.-U.Cam.

Whitehouse, 1939

M.Cam.-U.Cam.

KINGSTONIIDAE

Kobayashi,

933

M.Cam.-U.Cam.

ANOMOCARIDAE

Poulsen,

927

M.Cam.-U.Cam.

Raymond,

924

M.Cam.

Asaph iscinae Raymond,

924

M.Cam. (Asaphiscus)U.Cam.

Blountiinae Lochman,

944

M.Cam.-U.Cam.

BURLINGIIDAE

Walcott,

908

M.Cam.-U.Cam.

MENOMONIIDAE

Walcott,

916

ASAPHISCIDAE

MARJUMIIDAE

Kobayashi, .935

SOLENOPLEURIDAE

M.Cam. (Bolaspidella)U.Cam. M.Cam. (Modocia)U.Cam.

Angelin, 1854 Saoinae Hupe, 1953

Solenopleurinae Angelin, 1854

M.Cam. (Sao) M.Cam.

(Parasolenopleura)U.Cam. Solenopleuropsinae Thoral, 1947 Acrocephalitinae Hupe, 1953 Hystricurinae Hupe, 1953

M.Cam. (Badulesia) M.Cam.-U.Cam. U.Cam.

Kobayashi, 1935

M.Cam.-U.Cam.

LEIOSTEGIIDAE

Bradley, 1925

M.Cam.-U.Cam.

DAMESELLIDAE

Kobayashi, 1935

M.Cam.-U.Cam.

AVONINIDAE

Lochman, 1936

M.Cam.-U.Cam.

EOACIDASPIDIDAE

Poletaeva, 1957

M.Cam.-U.Cam.

CERATOPYGIDAE

Linnarson, 1869

M.Cam.-U.Cam.

KOMASPIDIDAE

Kobayashi, 1935

M.Cam.-U.Cam.

SHUMARDIIDAE

Lake, 1907

M.Cam.-U.Cam.

PAGODIIDAE

AVONINIDAE

Lochman, 1936

U.Cam.

EOACIDASPIDIDAE

Poletaeva, 1957

U.Cam.

Cambrian

Families

CERATOPYGIDAE

Linnarson, 1869

U.Cam.

KOMASPIDIDAE

Kobayashi, 1935

U.Cam.

SHUMARDIIDAE

Lake, 1907

U.Cam.

Palmer, 1954

U.Cam.

Lochman, 1956

U.Cam.

DIKELOCEPHALIDAE

Miller, 1889

U.Cam.

PTEROCEPHALIIDAE

Kobayashi, 1935

U.Cam.

Hupe, 1953

U.Cam.

Vogdes, 1890

U.Cam.

Hupe, 1953

U.Cam.

Lochman, 1956

U.Cam.

LONCHOCEPHALIDAE

Hupe, 1953

U.Cam.

DOKIMOCEPHALIDAE

Kobayashi, 1935

U.Cam.

CATILLICEPHALIDAE

Raymond, 1938

U.Cam.

ELVINIIDAE

Kobayashi, 1935

U.Cam.

TRICREPICEPHALIDAE IDAHOIDAE

HOUSIIDAE ILLAENURIDAE SHIRAKIELLIDAE PARABOLINOIDIDAE

RAYMONDINIDAE

•

Raymondininae Clark, 1924

U.Cam.

Cedariinae Raymond, 1937

U.Cam.

Llanoaspidinae Lochman, 1944

U.Cam.

COOSELLIDAE CHEILOCEPHALIDAE KAOLISHANIIDAE

Atlas o f Trilobite Photographs

Palmer, 1954

U.Cam.

Shaw, 1956

U.Cam.

Kobayashi, 1935

Kaolishaninae Kobayashi, 1935

U.Cam.

Manuyi nae Hupe, 1955

U.Cam.

Tingocephal nae Hupe, 1955

U.Cam.

PTYCHASPIDIDAE

118

Clark, 1924

Raymond, 1924

U.Cam.

SAUKIIDAE

Ulrich and Resser, 1933

U.Cam.

Hupe, 1955

U.Cam.

Rasetti, 1959

U.Cam.

Raymond, 1924

U.Cam.

Lu, 1954

U.Cam.

Kobayashi, 1933

U.Cam.

NORWOODIDAE

Walcott, 1916

U.Cam.

LECANOPYGIDAE

Lochman, 1953

U.Cam.

PLETHOPELTIDAE

Raymond, 1924

U.Cam.

Whittington, 1950

U.Cam.

ENTOMASPIDIDAE Ulrich in Bridge, 1930

U.Cam.

EUREKIIDAE LOGANELLIDAE HUNGAIIDAE DICERATOCEPHALIDAE TSINANIIDAE

HARPIDIDAE

OLENIDAE

Burmeister, 1843 Oleninae Burmeister, 1843

U.Cam. (Olenus) —i

Leptoplastinae Angelin, 1854

U.Cam.—>

(?) Pelturinae Hawle and Corda 1847

U.Cam.—>

(?) Triarthrinae Ulrich, 1930

U.Cam.->

RHODONASPIDINAE

Opik, 1963

U.Cam.

REMOPLEURIDIDAE Hawle and Corda, 1847 Richardsonellinae Raymond, 1924 ASAPHIDAE

U.Cam.—>

Burmaeister, 1843 Niobinae Janusson, 1959

U.Cam.

Promegalaspidinae Janusson, 1959

U.Cam.

(Niobella)-

(Promegalaspides) CONDYLOPYGIDAE

Raymond, 1913

L.Cam.-M.Cam.

EODISCIDAE

Raymond, 1913

L.Cam.-M.Cam.

PAGETIIDAE

Kobayashi, 1935

LCam.-M.Cam.

Cambrian

Families

AGNOSTIDAE

M'Coy, 1849 Agnostinae M'Coy, 1849

Ptychagnostinae Kobayashi, 1939

M.Cam.

Quadragnostinae Howell, 1935

M.Cam.

SPINAGNOSTIDAE

Howell, 1935

DIPLAGNOSTIDAE

Whitehouse, 1936 M.Cam.

Oidalagnostinae Opik, 1967

M.Cam.

Tomagnostinae Kobayashi, 1940

M.Cam.

Ammagnostinae Opik, 1967

M.Cam.

Pseudagnostinae Whitehouse, 1936

U.Cam.—>

Glyptagnostinae Whitehouse, 1936

U.Cam.

Aspidagnostinae Pokrovskaja, 1960 Howell, 1935

PHALACROMIDAE Hawle and Corda, 1847

Atlas of Ttilobitc Photographs

(Tomagnostus)

(Glyptagnostus)

Howell, 1937

Clavagnostinae Howell, 1937

MICRAGNOSTIDAE

(Ptychagnostus)

L.Cam.-M.Cam. (Baltagnostus, Peronopsis)U.Cam.

Diplagnostinae Whitehouse, 1936

CLAVAGNOSTIDAE

120

M.Cam.-U.Cam. (Homagnostus) —>

M.Cam.-U.Cam. U.Cam. U.Cam. —> M.Cam.-U.Cam.—>

Family Olenellidac Vogdes, 1893

PLATE

52

O L E N E L L U S Billings,

1861

(= Fremontia Raw, 1 9 3 6 ; Mesonacis W a l c o t t ,

1885;

Paedumias W a l c o t t ,

1910),

Olenellus fremonti W a l c o t t (x7.2). Lower Cambrian of British

Olenellus

Columbia.

Bonnia-

assemblage zone in

the St. Piran S a n d s t o n e (Peyto Limestone M e m b e r ) . (Gift o f J . R . Evans t o U C W M ; loaned b y F M N H ; id. RLS.) T h e classification of the Olenellidae has undergone extensive revision in recent years (Fritz 1 9 7 2 ; Bergstrom 1 9 7 3 b ) . Having changed generic affiliation several times,

fremonti

Olenellus

is presently reinstated

within the original denomination given by W a l c o t t , 1 9 1 0 . This trilobite typically has fourteen thoracic segments. Note exaggerated development of the pleural lobe of the third thoracic segment, said to be

macropleural

The

irregular

fractute along the axis is probably due to compression. T h e axial spine is hollow.

Cambrian Families

121

PLATE

53

Juvenile form o f

fremonti

Walcott

Olenellus (xlO),

from

the same sample which yielded the preceding example (RLS coll., id.; now at F M N H . ) In (a) the specimen, whitened with magnesium oxide, is photographed by standard technique, while in (b) the specimen is photographed while immersed in xylene. The optical contact of this medium (index ot refraction n = 1.5) with the surface crystalline layers of the sample provides a m u c h better representation of the anatomical features than in (a). Note that the axial spine is as long as the entire carapace.

122

Atlas o f Trilobite Photographs

Cambrian

Families

123

PLATE

54

Exuviae of a large

specimen o f Olenellus fremonti Walcott (x2) from the Lower C a m b r i a n Latham Shale of the Marble Mountains, San Bernardino County, California (RLS coll., id.). As a c o m m o n l y encountered o u t c o m e of the molting process, the cranidium slipped backward, overlapping the first two thoracic segments. T h e genal, pleural, and axial spines of the trilobites from this locality are often replaced by limonite and appear in brightly colored hues of yellow, red, and black, as is the case in this example.

124

Atlas o f Trilobite Photographs

PLATE

55

External impression of the exuviae of the trilobite

Olenellus clarki

(Resser) ( x 2 . 8 ) , also from the L a t h a m Shale o f the M a r b l e M o u n t a i n s ( R L S coll., id., n o w a t F M N H ) . This trilobite was previously assigned to

the genus Walcott,

Paedumias

1 9 1 0 (see

synonyms mentioned for plate 5 2 ) a n d bears great resemblance to

Olenellus

transitans

(Walcott) of the Atlantic faunal province. N o t e w i d e frontal area anterior to glabella, marked by median ridge, and third segment macropleurae. T h e latter, however, are not as wide as in O.

fremonti

(Walcott).

T h e opisthothorax (the posterior extension of the thorax) is partially exposed in this graceful example; the presence of this extension of the rachis is characteristic of the Olenellidae.

Cambrian

Families

125

PLATE

56

Particularly well preserved, although partially disarticulated, exuviae of another

specimen o f Olenellus

clarki

(Resser) from

the Latham Shale (x4.7)(RLS coll., id.). T h e only functional sutures in the cranidium of this ttilobite are those that rerain the visual surface. This is genetally released during molting or after death of the individual, leaving a gap between the palpebral lobe and the librigena, as is cleatly occutring in this example. A peculiar fractute and displacement o f part o f the macropleural third tetgite causes an apparent misalignment of the right and left side of the latter. Axial nodes are present on most axial rings. T h e axial spine and opistothorax are missing.

126

Atlas o f Trilobite Photographs

PLATE

57

Olenellus clarki

(Resser)

(x6.3). Lower C a m b r i a n o f Pioche, N e v a d a . Pioche Shale, D M e m b e r . Specimen collected by A f t o n Fawcett d u r i n g a field trip w i t h the a u t h o r ( R L S coll., id., n o w at F M N H ) . Part of a slab c o n t a i n i n g four partially c o m p l e t e individuals. S p e c i m e n whitened with magnesium oxide. T h e eyes of this individual are s o m e w h a t longer than the n o r m .

Cambrian

Families

PLATE

58

This somewhat stretched, yet captivating example

of OUnellus clarki (Resser) originates from the Nopah range of Southern California (x4.1). Specimen collected by Thomas Johnson (RLS coll., id.)

V

128

Atlas o f Trilobite Photographs

PLATE

59

Olenellus

mohavensis

(Resser) ( x 7 . 3 ) , a n o t h e r olenellid species f o u n d in the L a t h a m S h a l e o f the Marble Mountains of Southern

California

( R L S coll., id.). A disproportionate swollen

macropleura

of the third thoracic tergite a n d an advanced genal angle characterize this trilobite at all stages o f growth.

Cambrian

Families

129

PLATE

60

T h e characters noted in the preceding example become even more pronounced in

Bristolia bristolensis

Resser (x3.9)

from

the same Lower C a m b r i a n locality. External imptession (RLS coll., id.).

130

Atlas of Trilobite Photographs

PLATE

61

species found in the L o w e r

(Resser)(b)(x2.5), Bristolia bristolensis Resser (c)(x2.0),

C a m b r i a n Latham Shale o f

and

the M a r b l e M o u n t a i n s o f

(d)(x3.0). All of these species

S o u t h e r n California (RLS

have the glabella extending

C r a n i d i a of several olenellid

Bristolia insolens

Resser

coll., id.). A n evolutionary

to the frontal rim of the

trend t o w a r d a forward shift

cranidium. By contrast, the

of the genal angle is noted in

long-eyed

the sequence of short-eyed

clarki

olenellids that starts w i t h

its w i d e frontal b r i m , does

Olenellus jremonti

n o t seem to belong w i t h

(Walcott)

(a)(x2.3) and includes

Olenellus

form

Olenellus

(Resser)(e)(xl.8), w i t h

this sequence.

mohavensis Cambrian

Families

131

Family Holmiidae Hupe, 1953

PLATE

62

ANDALUSIANA

Sdzuy,

1 9 6 1 . Andalusiana sp. ( x l . 3 ) . Lower Cambrian, Djbel Ougrat, Morocco. (RLS coll., id.) Complete specimens of this trilobite have appeared in the fossil market of recent years in great number, unfortunately mostly defaced by poor preparation or arbitrary reconstruction. T h e granulated, reticulated surface of the carapace betrays the genuine portions of the specimen figured here, altered only by m i n o r repair. T h e genus Andalusiana,

indigenous

of Spain, as the n a m e implies, is k n o w n from the type

Andalusiana

comuta

Sdzuy. T h e M o r o c c a n species bears m a n y similarities w i t h the latter, w h i l e differing in some aspects. T h e palpebral lobes are here about one-third the length of the glabella, versus o n e - h a l f for

Andalusiana

cornuta.

Noticeable also is the falcate appearance of the broad genal spines, even m o r e p r o n o u n c e d than in the related genus Kjerulfia. T h e thorax contains sixteen segments. T h e axial notches or spines b e c o m e progressively m o r e pronounced, distally tapered and rectangular in cross section, toward the p y g i d i u m .

132

Atlas o f Trilobite Photographs

Family Wanneriidae Hupe, 1953

PLATE

63

Wanneria

walcottana

(Wanner), Lower Cambrian,

Kinzers

F o r m a t i o n , Lancaster, PA. (x2.8)(coIlected by M. Thomas, RLS coll.).

Although

r e m a r k a b l y well preserved in its undistorted entirety, this is an internal impression that does not show surface o r n a m e n t a t i o n or axial nodes that

s h o u l d be

present. Axial spine also n o t preserved.

Cambrian

Families

133

PLATE

64

Unusual assemblage of complete individuals

of Wanneria sp., from the Lower Cambrian of Sweden, at various stages of growth (x0.8). (Loaned through courtesy of Pio Pezzi.)

134

Atlas o f Trilobite Photographs

Family Redlichiidae Poulsen, 1927 Subfamily Redlichiinae Poulsen,

1927

PLATE

65

REDLICH1A Cossman,

1902.

Redlichia

forresti

(Etheridge J r . ) , Early Middle Cambrian, E m u Bay Shale near Big G u l l y , N o r t h Coast of Kangaroo Island, Australia, partially disarticulated exuviae ( x 4 . 2 ) . In this trilobite, facial sutures (opisthoparian) separating the glabella from the librigenae appear for the first time. (Photographed by the author at S M N H , through courtesy of Jan Bergstrom.)

Cambrian

Families

135

Family Dolcrolenidae Kobayashi, 1951

PLATE

66

DOLEROLENUS Leanza, 1 9 4 9 .

Dolerolenus

zoppii

(Meneghini) (x6.8) U p . L.Cambrian, Nebida Formation, Punta Manna member, Porto di Canalgrande, Sardinia, Italy. In this specimen, the exoskeleton is still replaced by a hard film of hematite, which is m o r e generally altered to limonite. (Photo courtesy of Franco Todde, reproduced by permission.)

136

Atlas o f Trilobite Photographs

P LAI £ 6 7 Dolerolenus

zoppii

(Meneghini) (x7.4), from the same locality as preceding plate. H e r e m o s t o f the original h e m a t i t i c replacement o f the exoskeleton is altered to a bright y e l l o w film o f l i m o n i t e . External impression. (Loaned through courtesy o f Bruno Corti.)

Cambrian

Families

137

Family EUipsocephalidae Matthew, 1887 Subfamily Ellipsocephalinae Matthew,

138

1887

Atlas of Trilobite Photographs

PLATE

68

E L L I P S O C E P H A L U S Zenker,

Ellipsocepbalus boffi

(Schlotheim)

1833. (xl.8).

O k l a h o m a . ) This trilobite c o m m o n l y occurs in densely populated assemblages,

M . C a m b r i a n , Jinetz, Bohemia. (Loaned

k n o w n since the early nineteenth

by Geological Enterprises, A r d m o r e ,

century.

Family Conocoryphidae Angelin, 1854

PLATE

69

CONOCORYPHE and C o r d a ,

Conocoryphe

Hawle

1847.

sulzeri

Schlotheim ( x l . 8 ) . M . C a m b r i a n (Jinetz Shale), Beroun near Prague, Bohemia. (Loaned b y M C Z . ) This trilobite was blind. T h e specimen on the l o w e r lefthand corner is a negative mold, the o t h e r t w o are positive casts (steinkerns).

Cambrian Families

139

PLATE

70

Conocoryphe

sulzeri

Schlotheim, as for preceding plate (x3.4). (RLS coll.; courtesy of M C Z , now at F M N H . ) W h a t blackand-white photography cannot convey is, of course, the color of some of the trilobite specimens. In this case, the trilobite is coated by a bright yellow-ochre film of limonite that contrasts with the tan matrix background.

140

Atlas o f Trilobite Photographs

PLATE

71

BAILIELLA

Matthew,

1 8 8 5 . Bailielk manuelensis (Hutchinson), M. Cambrian,

Eccaparadoxides

bennetti

beds, Kellygrews, Conception

Bay,

Newfoundland

(x3.6).

Blind trilobite. ( R L S coll., id.)

Cambrian

Families

141

142

Atlas o f Trilobite Photographs

PLATE

72

C T E N O C E P H A L U S Hawle and C o r d a , 1847. Resser,

Cranidium M.

of

Cambrian,

Ctenocephalus howelli Paradoxides davidis

beds, east side of M a n u e l s River, Manuels, N e w f o u n d l a n d ( x 5 . 9 ) ( R L S coll., id.). This blind trilobite is closely related to

Conocoryphe,

with

fifteen

thoracic

segments instead of fourteen. T h e location of the preglabellar boss drifts toward the anterior border in other species. T h e granulated surface of the convex parts of the cranidium is particularly noticeable and well preserved in this example.

Cambrian

Families

143

Family Zacanthoididae Swinnerton,

PLATE

1915

73

ZACANTHOIDES Walcott, 1888.

Zacanthoides

typicalis

Walcott ( x l l ) . Chisholm Shale, M. Cambrian, Half M o o n M i n e , Pioche, Nevada. (Specimen collected by Afton Fawcett. RLS coll., id., n o w at FMNH.) O n c e again it is worth showing the comparison berween standard photography o f the specimen in air (a) and what can be seen with xylene i m m e r s i o n (b). T h e long axial spine originates from the eigth thoracic axial ring.

144

Atlas of Trilobite Photographs

Cambrian Families

145

PLATE

74

Zacanthoides

typicalis

W a l c o t t ( x l 5 . 8 ) . Same origin as for preceding plate. (Specimen collected by Afton Fawcett. RLS coll., id., now at FMNH.) T h e trilobite exoskeleton is partly preserved on both sides of the bedding plane. W i t h the aid of xylene immersion, the parts which are missing in (b) can be identified in (a), so that a c o m p l e t e reconstruction can be obtained.

146

A t l a s of T r i l o b i t e P h o t o g r a p h s

Cambrian Families

147

Family Dorypygidae Kobayashi, 1935

PLATE

75

O L E N O I D E S Meek,

servants

148

Atlas o f Trilobite Photographs

1877.

Olenoides

R o m i n g e r ( x l . 8 ) . Burgess Shale,

standard technique; and (b) while immersed in xylene. N o t e the power o

Stephen F o r m a t i o n , M . C a m b r i a n ,

the second m e t h o d in revealing details

British C o l u m b i a . (Loaned by M C Z . )

the c o n t o u r of the carapace. T h e

P h o t o g r a p h y of the same pair of

trilobites from this famous locality are

trilobites by t w o approaches: (a) by

often found w i t h soft parts preserved.

Unfortunately, this specimen does n o t show the appendages often seen emerging from the edge of the carapace. Compression of the glabella outlines the underlying hypostoma.

Cambrian Families

149

PLATE

76

A striking example of

Olenoides superbus Walcott from the Marjum Formation, M. Cambrian, east side of Wheeler Amphitheatre, House Range, Utah ( x l . 8 ) . (Photographed by the author, through courtesy of Eddie Cole, who collected and prepared the specimen.)

150

Family 1916

Dolichometopidae

Walcott,

PLATE

77

B A T H Y U R I S C U S Meek,

Association of

1873.

Batbyuriscus fimbriatus

Robison (right a n d left) with laevinucha Robison

(center)

Modocia

(x2.7).

between M a r j u m pass and A n t e l o p e Springs, M i l l a r d C o u n t y , Utah. (Collected by A f t o n Fawcett. R L S coll id., n o w a t F M N H . )

M a r j u m Formation, M. Cambrian,

Cambrian Families

151

PLATE Bathyuriscus

78 fimbriatus

Robison (x6.3). Same origin as previous plate. (Collected by Afton Fawcett. RLS coll., id., now at F M N H ) . O n e o f the missing libriginae overlaps, over-turned, the right-hand side of the thorax, showing prominent radiating alimentary diverticula (prosopon).

152

Atlas of Trilobite Photographs

PLATE

79

A complete exoskeleton o f

Bathyuricus

fimbriatus

Robison ( x l 2 ) . Same origin as previous plate (collected by A f t o n Fawcett, R L S coll., id., now at F M N H ) .

Cambrian

Families

153

PLATE

80

HEMIRHODON Raymond,

1937.

Hemirhodon

amplipyge

Robison (x2.1). M a r j u m Formation, M a r j u m Pass, Millard County, Utah. (Collected by Alton Fawcett. RLS coll., id., now at FMNH.) A beautiful example of this large trilobite.

154

Atlas o f Trilobite Photographs

Family Ogygopsidae Rasetti, 1951

PLATE

81

O G Y G O P S I S Walcott,

Ogygopsis klotzi

1889.

(Rominger),

M . C a m b r i a n , British C o l u m b i a ( x 2 . 0 ) . (Latex m o l d o f specimen loaned from MCZ.)

Cambrian Families

155

Family Paradoxididae Hawle and Corda, 1847 Subfamily Paradoxidinae Hawle and C o r d a 1 8 4 7

PLATE

82

PARADOXIDES Brongniart,

1822.

Paradoxides paradoxissimus (Wahlenberg), M. Cambrian, Oltorp, Vastergotland, Sweden ( x l . 4 ) . This is the type specimen, preserved in black alum shale, described in Westergard ( 1 9 5 2 ) , pi. 8, fig. 2, originally

Paradoxides

tessini Brongniart ( S M N H No. A r 4 7 4 8 7 ) . T h e upper specimen is an external impression (photographed by the author a t S M N H , through courtesy of Jan Bergstrom).

156

Atlas of Trilobite

PLATE Colorful

83 preservation

of the exuviae in another example of

Paradoxides paradoxissimus ( W a h l e n b e r g ) , from the same locality as plate 8 2 ( x 0 . 8 4 ) . ( P h o t o g r a p h e d by the a u t h o r at the Paleontological Institute, U n i v e r s i t y o f Uppsala, S w e d e n . )

Cambrian

Families

157

PLATE

84

D e v e l o p m e n t a l stages

in all growth stages of

suggest some kind of

(late meraspid degrees)

Olenellus (see plates

size sorting in molting.

in the Middle Cambrian

5 2 - 6 0 ) . In the early

P. gracilis

stages of g r o w t h as in

indistinguishable from

trilobite

gracilis

Paradoxides

(Boeck), M .

C a m b r i a n , from Jince

is practically

the examples s h o w n in

P.

this plate, the eye lobe

except for the n u m b e r

paradoxissimus,

of segments in adult

(Jinetz), Bohemia. The

extends over a larger

specimen in (a) (x8.6)

portion of the cranidial

individuals, twenty in

exhibits much elongated

length than in the

the Bohemian (see plate

genal spines as well as

fully g r o w n individuals

1), t w e n t y - o n e in the

pleural spines of the

such as that of plate 1.

Swedish

second thoracic segment.

O t h e r late meraspid

Specimens in (a) and

T h e latter character is

exuviae are s h o w n in

(b) w h i t e n e d

c o m m o n to meraspid

(b) ( x l . 9 ) in an

magnesium oxide.

stages of other

assemblage o f o n e

(RLS coll., donated by

paradoxidids (see plate

internal m o l d a n d t w o

D r . Petr Storch of the

Paradoxides. with

9 2 ) and is reminiscent

external impressions of

Geological Institute,

of the long pleural

approximately equal

A c a d e m y o f Sciences

spines associated with

size. T h i s and o t h e r

of the Czech Republic,

the third segment

similar assemblages

Prague.)

macropleurae observed

from the same layers

158

PLATE

85

A well-preserved, d i m i n u t i v e

example of Paradoxides gracilis (Boeck)(x4.5). M. Cambrian, Jinetz, B o h e m i a . ( R L S coll.; courtesy o f M C Z , n o w a t F M N H . ) Specimen whitened with magnesium oxide.

Cambrian Families

159

PLATE

86

Beautifully preserved specimen

of Paradoxides davidis davidis Salter, M. Cambrian, west side of Manuels River, Newfoundland ( x l . 3 ) . This trilobite is first known from St. David's, Wales. A complete account of the evolutionary sequence that transpires from its fossil record in the Avalon Peninsula of N e w f o u n d l a n d is given in A p p e n d i x A of chapter 3. This particular example represents o n e of the earliest appearances of this trilobite at this locality. (External impression, specimen loaned t h r o u g h courtesy o f

160

M a r k Utlaut.)

Atlas o f Trilobite Photographs

PLATE

87

ACADOPARADOXIDES Snajdr,

1958.

Acadoparadoxides sp.,

M.

C a m b r i a n , Sidi A b d a l l a h ben el H a d j , M o r o c c o (xO.65) ( R L S coll.). T h e s e giant trilobites (up to 50 cm in length) are extracted commercially in staggering n u m b e r and generally dubiously restored. T h e example s h o w n , w i t h eighteen thoracic segments, corresponds closely to

Acadoparadoxides

harlani

(Green), as described by W a l c o t t ( 1 8 8 4 ) , pi. IX., a n d may well represent the same trilobite. This w o u l d not be surprising since, as s h o w n in figure 1 8 , Maritime N o r t h America faced what was to become the Anti-Atlas Range in the M i d d l e C a m b r i a n .

Cambrian

Families

161

PLATE

88

HYDROCEPHALUS Barrande,

hicksi

1846.

Hydrocephalus

(Salter), M. Cambrian,

west side of Manuels River, Manuels, Newfoundland (x2.5). The libriginae are missing. Latex mold from external impression. (RLS id.)

162

Atlas of Trilobite Photographs

PLATE

89

ECCAPARADOXIDES Snajdr,

1958.

Eccaparadoxides eteminicus

(Matthew),

M. C a m b r i a n , E.

bennetti

beds,

Kellygrews, C o n c e p t i o n Bay,

Newfoundland

( x l . 5 ) . The cranidium is i n c o m p l e t e a n d the libriginae are missing. (RLS coll., id.)

163

PLATE

90

Eccaparadoxides

pusillus

(Barrande), M. Cambrian

Paradoxides davidis

beds at

Kellygrews, Conception Bay, Newfoundland (x3.6) ( R L S coll., id.) The name of this long-eyed Paradoxidid is n o w regarded (Snajdr 1 9 5 8 ) as synonymous with

Paradoxides

rugulosus Corda.

164

Atlas of Trilobite Photographs

PLATE

91

Eccaparadoxides

oelandicus

(Sjogren), M. C a m b r i a n E.

pinus

zone, B o r g h o l m ,

Oland, Sweden (x6.7). Type specimen of late meraspid stage, described by W e s t e r g a r d ( 1 9 3 6 ) , pi. 2 , f i g . 4. (SMNH No. Ar 4 6 2 4 4 , photographed by the a u t h o r a t S M N H , t h r o u g h courtesy of Jan Bergstrom.)

Cambrian Families

165

PLATE

92

Eccaparadoxides pinus (Holm), M. Cambrian, Borgholm, Oland, Sweden (x9). Type specimen of late meraspid stage, described by Westergard (1936), pi. V, fig. 7. (Photographed by the author at GSS, Uppsala, through courtesy of Christer Akerman.)

166

Atlas o f Trilobite Photographs

PLATE

93

Eccaparadoxides torelli ( H o l m ) , M . Cambrian, Borgholm, Oland, Sweden (x2.0). T y p e specimen, described by Westergard ( 1 9 3 6 ) , pi. VIII, f i g . l b .

(Photographed by the author at GSS, through courtesy of Christer Akerman.)

Cambrian Families

167

Subfamily Xystridurinae Whitehouse, 1939

PLATE

94

XYSTRIDURA Whitehouse, 1939. Xystridura saintsmithi (Chapman), M. Cambrian, Beetle Creek, Queensland, Australia (x4.1). ( S M N H No. Ar 55079, photographed by the author at S M N H , through courtesy of Jan Bergstrbm.)

168

Atlas o f Trilobite Photographs

Subfamily Centropleurinae Angelin, 1854

PLATE

95

A N O P O L E N U S Salter, 1864. Anopolenw henrici Salter, M. Cambrian, Paradoxides davidis beds, west side of Manuels River, Manuels, Newfoundland (x2.3). One of the released free cheeks, and its long, slender genal spine, overlaps the body. T h e surface of the carapace is finely textured. This trilobite was first discovered in Wales. (Gift by Jan M. Chabala, specimen whitened with magnesium oxide.)

Cambrian Families

169

Family Agraulidae Raymond, 1 9 1 3

PLATE

96

AGRAULOS Hawle and Corda, 1847. Agraulos socialis Billings (x6.9). M. Cambrian, Paradoxides hicksi beds, west side of Manuels River, Manuels, Conception Bay, Newfoundland (RLS coll., id.). The libriginae are missing. Specimen whitened with magnesium oxide.

170

Atlas of Trilobite Photographs

Family Alokistocaridae Resser, 1 9 3 9

PLATE

97

A L O K I S T O C A R E Lorenz, 1906. Alokistocare harrisi Robison (x6.5). Wheeler Shale, M. C a m b r i a n , Antelope Springs, Utah. (Collected by Afton Fawcett. RLS coll., id., now at F M N H . ) Following Opik (1961), the family Alokistocaridae Resser, 1939, is regarded as a younger synonym of Papyriaspididae Whitehouse, 1939. Specimen whitened with magnesium oxide.

Cambrian Families

171

172

Atlas o f Trilobite Photographs

PLATE

98

Alokistocare piochensis (Walcott) (x4.0). Chisholm Shale, M. Cambrian, Half Moon Mine at Pioche, Nevada. (Collected by Afton Fawcett. RLS coll., id., now at F M N H . ) One of the free cheeks is displaced.

Cambrian Families

173

PLATE

99

PIOCHASPIS, n.g. Piochaspis sellata, n. sp., holotype. Chisholm Shale, M. Cambrian, Half Moon Mine at Pioche, Nevada (xl4.4). (Collected by Afton Fawcett. RLS coll., id.) Plate 99 (a) shows the specimen whitened with magnesium oxide; plate 99 (b) shows the same specimen while immersed in xylene, revealing additional details. A morphometric study by the author of the population of ptychoparid trilobites from this locality (unpublished, but discussed with F. Rasetti and A. Palmer) shows this trilobite to be distinct from Alokistocare piochensis (Walcott) and Alokistocare packi (Resser) from the same beds. The description of this new trilobite is the following: Anterior border convex, anterior border furrow with median inbend, preglabellar field depressed mesially, elevated laterally; glabella tapering forward, rounded in

1 74

Atlas of Trilobite Photographs

from,

with

three pairs

of shallow lateral furrows; axial furrows deep, facial sutures diverging in front of eyes; eye ridges narrow, - 0.3 of glabellar width at midlength, palpebral areas convex, narrow anteriorly, posterior area of fixigena deeply furrowed. Librigenae with pointed genal spines, flaring posteriorly at - 45°. Thorax of seventeen segments; pleurae grooved, pleural spines short, outward pointing. Pygidium small, devoid of border, with prominent axis, one axial ring. The entire collection of 67 complete individuals relating to this study is now at Field Museum of Natural History. F M N H record numbers for the various groups are: Piochaspis sellata, holotype, PE 5 4 1 1 5 ; Piochaspis sellata, 48 syntypes, PE 5 4 1 1 6 54163; Alokistocare packi, 10 specimens, PE 5 4 1 6 4 - 5 4 1 7 3 ; Alokistocare piocbensis, 8 specimens, PE 54174-54181.

Cambrian Families

1 75

PLATE

100

ELRATHIA Walcott, 1924. Elrathia kingii (Meek)(xl.8). Wheeler Shale, Antelope Springs, Millard County, Utah. (Collected by Aiton Fawcett. RLS coll., now at FMNH.) In this

1 76

Atlas of Trilobite Photographs

fossilization, the trilobite exoskeleton is replaced by calcite; the crystals of this mineral extend in a so-called "cone-incone" structure well beneath the original carapace.

PLATE

101

Elrathia kingii (Meek) (x3.5). Same origin as for previous plate. (Collected by Afton Fawcett. RLS coll., now .it l-'MNI I.) Specimen of approximately maximum size for this species.

Cambrian Families

177

PLATE

102

Elrathia kingii (Meek) in association with the agnostid Peronopsis interstricta (White) (x.8). (Collected by Afton Fawcett. RLS coll., now at FMNH.) Various stages of growth of Elrathia are represented in this unusual plate.

178

Atlas of Trilobite Photographs

Family Ptychopariidae Matthew, 1 8 8 7 Subfamily Ptychopariinae Matthew, 1887

PLATE

103

PTYCHOPARIA Hawle and Corda, 1847. Ptychoparia striata (Emmrich), M. Cambrian, Jince Formation, Jince, Bohemia (x3.0). This genus is thought to be the progenitor of a very broad group of trilobites. In this example, two posterior thoracic segments and the

pygidium are missing (or enrolled underneath). T h e striated surface of the cranidium, origin of the adjectival name of this trilobite, is evident in this photograph. (RLS coll.)

179

Family Asaphiscidae Raymond, 1 9 2 4 Subfamily Asaphiscinae Raymond, 1924

PLATE

104

ASAPHISCUS Meek, 1873. Asaphiscus wheeleri Meek (x3.0). Wheeler Formation, M. Cambrian, Antelope Springs, Millard County, Utah. (Collected by Afton Fawcett. RLS coll., now at FMNH.) In this group, Asaphiscus is associated with Elrathia kingii.

180

Atlas o f Trilobite Photographs

PLATE

105

Asaphiscus wheeleri M e e k ( x l . 7 ) . Same origin as for preceding plate (RLS coll., now at F M N H . ) This is an extraordinarily large specimen. Photograph printed from color slide. T h e original specimen is black.

Cambrian

Families

181

Family Menomonidae Walcott, 1 9 1 6

PLATE

106

BOLASPIDELLA Resser, 1937. Bolaspidella housensis (Walcott) (x4.7). M. Cambrian, Wheeler Formation, Antelope Springs, Utah. (Collected by Afton Fawcett. RLS coll., id., now at FMNH.) Specimen whitened with magnesium oxide. On the uncoated specimen the trilobites appear black on a light gray soft matrix.

182

Atlas o f Trilobite Photographs

Family Marjumiidae Kobayashi, 1 9 3 5

PLATE

107

M O D O C I A Walcott, 1924. Modocia typicalis (Resser), M. Cambrian, M a r j u m Formation between M a r j u m Pass and Wheeler Amphitheatre, Millard County, Utah (x4.1). (Specimen collected by Robert Harris, RLS coll., id.)

Cambrian

Families

183

Family Solenopleuridae Angelin, 1 8 5 4 Subfamily Saoinae Hupe\ 1953

PLATE

108

SAO Barrande, 1846. Sao hirsuta Barrande, M. Cambrian, Jince Formation, Skrije, Bohemia (xl3.5) (RLS coll.) Barrande gave a detailed description of the ontogenic stages of this trilobite. Several stages are visible in this photograph, notably a larval stage (metaprotaspis) at the left of the largest exuviae (late meraspid), and an early meraspid degree at the upper right.

184

Atlas of Trilobite Photographs

Subfamily Solenopleurinae Angelin, 1854

P LAT E 1 0 9 PARASOLENOPLEURA Westetgard, 1953. Parasolenopleura aculeata (Angelin), M. Cambrian, Ptychagnostus gibbus zone, Kvarntorp, Narke, Sweden ( x l . 7 ) . (Photographed by the author at the Uppsala Paleontological Institute.)

Cambrian

Families

185

Subfamily Solenopleuropsinae Thoral, 1947

PLATE

110

SOLENOPLEUROPSIS Thoral, 1947. Solenopleuropsis variolaris (Salter), M. Cambrian, Paradoxides davidis beds, east side of Manuels River, Manuels, Newfoundland (x6.6) (RLS coll., id.). The prominent tuberculation of the convex surfaces of this trilobite is seen as a perforation pattern in this external impression. The pygidium is missing. Specimen whitened with magnesium oxide.

186

Atlas of Trilobite Photographs

PLATE

11 1

BADULESIA Sdzuy, 1967. Badulesia tenera (Hartt), M. Cambrian, Upper Paradoxides bennetti beds, Kellygrews, Conception Bay, Newfoundland. (x7.6). This trilobite is characteristic of the Middle Cambrian of the Asturias, Spain. Its appearance in the Avalon Peninsula of Newfoundland is sttiking and in accord with the continental drift map of figure 18, which places the Avalon and Spain on an almost contiguous shoreline east of Maritime North America. Specimen whitened with magnesium oxide. (RLS coll., id.)

Cambrian Families

187

Family Olenidae Burmeister, 1 8 4 3 Subfamily Oleninae Burmeister, 1843

PLATE

112

OLENUS Dalman, 1827. Olenus gibbosus (Whalenberg), M . - U . Cambrian, Outwoods Shale, Purley Quarry, Mancetter, Warwickshire (xlO). (Gift by Robert J. Kennedy, RLS coll.)

188

Atlas o f Trilobite Photographs

Family Asaphidae Burmeister, 1 8 4 3 Subfamily Niobinae Janusson, 1959

PLATE

113

NIOBELLA Reed, 1931. Niobella aurora Westergard, U. Cambrian, Olenus Shale, Raback, Kinnekulle, Vastergotland, Sweden (x3.0). This is the type specimen described by Westergard ( 1 9 3 9 ) in plate II, fig. 1. (Photographed by the author at C S S , Uppsala, through couttesy of Christer Akerman.)

Cambrian Families

189

Subfamily 1959

PLATE

Promegalaspidinae

Janusson,

114

PROMEGALASPIDES Westergard, 1939. Promegalaspides kinnekullensis Westergard, U. Cambrian, Olenus Shale, Raback, Kinnekulle, Vastergotland, Sweden (x3.2). T y p e specimen described by Westergard (1939) in plate I, fig. 1. (Photographed by the author at GSS, Uppsala, through courtesy of Christer Akerman.)

190

Atlas of Trilobite Photographs

Family Agnostidae M'Coy, 1 8 4 9 Subfamily Agnostinae M ' C o y , 1849

PLATE

115

H O M A G N O S T U S Howell, 1935. Homagnostus obesus (Belt), M . - U . Cambrian, Outwoods Shale, Purley Quarry, Mancetter, Warwickshire ( x l 4 ) . (Gift by Robert J. Kennedy, RLS coll.)

Cambrian Families

1 91

Subfamily Ptychagnostinae Kobayashi, 1939

PLATE

116

PTYCHAGNOSTUS Jaekel, 1909. Ptychagnostus richmondensis (Walcott) (xl6), Marjum Formation, M. Cambrian, mouth of Antelope Springs Canyon, Millard County, Utah. (Collected by Afton Fawcett. RLS coll., id., now at F M N H . ) Specimen whitened with magnesium oxide. The radiating ornamentation of the cephalon is attributed to alimentary diverticula.

192

Atlas of Trilobite Photographs

PLATE

117

Ptycbagnostus richmondemis (Walcott) (x6.1), Wheeler Shale, M. Cambrian, Antelope Springs, Millard County, Utah. (Collected by Afton Fawcett. RLS coll., id., now at F M N H . )

Several specimens show axial spine on second thoracic segment. T h e surface features emphasized in plate 116 are very difficult to observe without whitening.

Cambrian Families

193

PLATE

118

Ptychagnostus ciceroides (Matthew), M. Cambrian, Paradoxides davidis beds, Manuels, Newfoundland (x9). (RLS coll., id.)

194

Atlas of Trilobite Photographs

Family Spinagnostidae Howell, 1 9 3 5

PLATE

119

B A L T A G N O S T U S Lochman, 1944. Baltagnostus eurypyx Robison (x23). Wheeler Shale, M. C a m b r i a n , Antelope Springs, Millard County, Utah. (Collected by Afton Fawcett. RLS coll., id., now at F M N H . ) Note posterolateral marginal spines on pygidium.

Cambrian

Families

195

PLATE

120

PERONOPSIS Corda, 1847. Peronopsis interstricta (White) (x21). Wheeler Shale, M. Cambrian, Antelope Springs, Millard County, Utah. (RLS coll., id., now at F M N H . ) A much enlarged view, printed from a color slide.

196

Atlas of Trilobite Photographs

PLATE

121

Peronopsis interstricta (White) (x6.3). Same origin as preceding plate. (Collected by Afton Fawcett. RLS coll., id., now at F M N H . ) Assemblage of many individuals.

Cambrian Families

197

Family Diplagnostidae Whitehouse, 1 9 3 6 Subfamily Diplagnostinae Whitehouse, 1936

PLATE

122

T O M A G N O S T U S Howell, 1935. Tomagnostus perrugatus (Groenwall)(x9). M. Cambrian, lowest Paradoxides davidis beds, west side of Manuels River, Manuels, Newfoundland. (Loaned through courtesy of Mark Utlaut.)

1 98

Atlas of Trilobite Photographs

PLATE

123

Tomagnostus perrugatm (Groenwall), same origin as preceding plate (x5). (RLS coll., id.) This is a small portion of a large plate covered with the

exuviae of this agnostid, abundant at the base of the P. davidis beds, infrequent at other depths.

Cambrian Families

1 99

Subfamily Glyptagnostinae Whitehouse, 1936

PLATE

124

GLYPTAGNOSTUS Whitehouse, 1936. Glyptagnostus sp., M . - U . Cambrian, Ourwoods Shale, Purley Quarry, Mancetter, Warwickshire (xl3.5). (Gift of R. J. Kennedy, RLS coll.)

200

Atlas of Trilobite Photographs

4.2

Ordovician Families

AGNOSTIDAE

M'Coy, 1849 Agnostinae M'Coy, 1849