EARTH
SCIENCES
LIBRARY

AN
INTRODUCTION TO PALAEONTOLOGY

AMAUROCERAS FERRUGINEUM (SIMPSON), AN AMALTHEID FROM

THE DOMERIAN OF WHITBY.

Photographs by J. W. Tutcher of the holotype in Whitby Museum, repro-
duced from "Type Ammonites," by S. S. Buckman, vol. Hi. (1919).

a, Side view. (X§.) Radial line painted black, b, Peripheral (ipertural)
view. (Xi. ) c, Side view. (X2.) Suture-lines strengthened on
negative, guide-line in white. d, Side view. (X2.) Intervals
between certain sutures painted on print, e, Right half of suture-
line, straightened out. (X4.)

AN 'INTRODUCTION

TO

PALEONTOLOGY

A. MORLEY DAVIES

D.Sc., A.R.C.S., F.G.S.

HONORARY FELLOW OF THE ROYAL GEOGRAPHICAL SOCIETY J LECTURER IN
PALAEONTOLOGY, IMPERIAL COLLEGE OF SCIENCE AND TECHNOLOGY

LONDON

THOMAS MURBY & CO.

i, FLEET LANE, LUDGATE CIRCUS, E.G. 4
1920

0 E.

EARTH
SCONCES
LIBRARY

PREFACE

MY object in writing this book has been to introduce
the principal facts, problems, and results of Palaeontology
to those who have to study it as a self-contained science,
whether students of Geology or amateur fossil collectors.
Though, from an academic point of view, Palaeontology
cannot claim the status even of a branch of Biology,
being an aggregate of accidentally detached fragments of
that science ; yet, from the practical point of view, its
special technique and applications give it an individuality
as worthy of recognition as that, for instance, of Astro-
physics. Valuable as a preliminary training in general
Biology is to the student of Palaeontology, I have tried to
show that the latter, even when studied by itself, need
not be reduced to a mere descriptive catalogue of fossils.
My ideal (not always attained) has been, in treating
each great group of fossils, first to describe with some
fulness a few common species from which an idea of
the general characters and range of variation may be
obtained ; and then to give a brief systematic account of
the group. The reader should not be satisfied unless he
actually handles the selected species and compares them,
point by point, with the description. When he cannot
obtain the actual species described, he should take the
nearest allied form accessible, and determine precisely
how it differs from that described. Only by such
definite practical work can he begin to qualify himself to
become a palaeontologist.

495964

vi PREFACE

The writing of this book was begun in 1913, and
interrupted for nearly four years by pressure of war-
work. It was at first intended that it should be
illustrated by original drawings of fossils, and a certain
number of such drawings were made before the war
from specimens in the collection of the Imperial College.
I have to record my grateful thanks to the Governors
of the Imperial College and Professor Watts, F.R.S.,
for permission for the use of these specimens, and to
Mr. F. G. Percival, B.Sc., the artist. Subsequently it
became evident that to continue this method of illustra-
tion would delay the work too much, and that for the
purposes of the book diagrammatic copies of figures in
standard monographs would be equally useful. I have
to thank Miss L. Lyle, F.L.S., and my wife for these
later figures, about four-fifths of the total.

Mr. S. S. Buckman has been so kind as to read the
manuscript of the chapters on Brachiopoda and Cepha-
lopoda, and I have profited by his criticisms and
suggestions, and incorporated additional information
to a much greater extent than it was practicable to
acknowledge in the text. I cannot sufficiently express
my gratitude to him for this help, as well as for the
generous loan of the block which forms the frontis-
piece— a beautiful example of Mr. Tutcher's photographic
skill, from the latest part of Mr. Buckman's " Type

Ammonites."

A. MORLEY DAVIES.

February 2, 1920.

CONTENTS

PAGE

PREFACE

I. THE BRACHIOPODA .- r

II. THE LAMELLIBRANCHIA 45

III. THE GASTROPODA - 94

IV. THE CEPHALOPODA - - I23
V. THE TRILOBITA AND OTHER ARTHROPODA - - 191

VI. THE VERTEBRATA - - 221

VII. THE ECHINODERMATA - 24°

VIII. THE GRAPTOLITES AND CORALS - - 281

IX. THE PORIFERA AND PROTOZOA - 3l6

X. THE VEGETABLE KINGDOM - - 331

XI. THE COLLECTION AND PRESERVATION OF FOSSILS- 345

XII. THE RULES OF NOMENCLATURE - 354

APPENDIX I. THE DIVISIONS OF GEOLOGICAL TIME - 377

APPENDIX II. STRATIGRAPHICAL PALEONTOLOGY - - 399

INDEX - - - - - - - 4°7

LIST OF ILLUSTRATIONS

1. Terebratula intermedia -

2. Ornithella obovata 3

3. Brachial Loops - -6

4. Hemiihyris psittacea - X3

5. Bathonian Terebratulacea 21

6. Leptccna rhomboidalis - -26

7. Various Brachiopods 3°

8. Palaeozoic Rhynchonellids 36

9. Deltidial Plates of Rhynchonellids 37

10. Evolution of the Loop in Long-looped Terebratulacea 40

11. Pcctunculus glycimeris - - 48

12. Pcctunculus glycimeris - - 49

13. Nucula - 53

14. Trigonia bronni - 55

15. Cyrena - - 57

16. Heterodont Hinges - 59

17. Crassatella ponder osa - - 62
1 8. Meretrix incrassata - 63

19. Pseudomonotis ecliinata - • 65

20. Ostrea ventilabrum ~ ^7

21. Corbula gallica - - 7°

22. Pholadomya - 71

23. Lamellibranchs - 7&

24. Heterodont Lamellibranchs - 83

25. Fixed and Burrowing Lamellibranchs - 87

26. Emarginula fissura - 9&

27. Turritella imbricataria - 97

28. Natica multipunctala - - IO°

29. Cerithium scrratum - 102

LIST OF ILLUSTRATIONS

FIG.

30. Rimella fissurella . 103

31. Trivia avellana - - 105

32. Fusus porrcctus - - 107

33. Planorbis discus - - 108

34. Gastropods - . - 114

35. Gastropods _ 118

36. Orthoceras - - - 125

37. Nautilus albcnsis - - 127

38. Nautilus excavatus • - - 128

39. Asteroceras obtusurn - - - - 131

40. Tetramcroceras obovatum - - 137

41. Cephalopod Coiling Cycle - 141

42. Manticoceras aff. retrorsiun - - 152

43. Manticoceras rliyncJiostoma - 153

44. Gastrioceras carbonari u in - 155

45. Apertural Margins of Cephalopoda - - 156

46. Apertural Margins of Cephalopoda - - 158

47. Septal Sutures of Early Ammonoidea - - 162

48. Septal Sutures of Jurassic and Cretaceous Ammonoidea 164

49. Whorl-shape in Ammonoidea - - 166

50. Lunuloceras bright i - - 176

51. Olcostephanus astieriaiius - 178

52. Mortoniceras inflatum - - 180

53. Belemnites - - 185

54. Belemnites - 188

55. Calymene blumenbachi - - 195

56. Dalmanites caudatus - - 199

57. Triarthrus becki - - - 201

58. Trinucleus concentricus - - 203

59. Trilobites - 208

60. Cheirurus bimucronatus - 210

61. Hypostomes of Various Trilobites - 211

62. Lanarkia spinosa - 223

63. Fish Teeth and Scales - - 224

64. Tooth Structure - - 226

65. Caudal Fins - 228

66. Fins and Limbs - - . - 232

LIST OF ILLUSTRATIONS xi

FIG. PAGE

67. Cnpressocrinus - - 245

68. Amphoracrinus atlas - - 247

69. Marsupites testudinarius - 249

70. Crinoid columnals - 251

71. Crinoidea - - 252

72. Echinosphcera aurantiuni - 256

73. Cystoidea - - 257

74. Pentremites pyriformis - - 259

75. Hemicidaris intermedia - - 262

76. Conulus albogalerus - - 265

77. Micr aster cor-anguinum - 267

78. Regular Echinoids - - 272

79. Irregular Echinoids - - 274

80. Collyrites ringens - 275

81. Echinocorys scutatus - - 277

82. Holaster subglobosus - - 278

83. Didymograptus murchisoni - 282

84. Didymograptus minutus - 283

85. Evolution of the Dichograptidae - 285

86. Climacograptus wilsoni - 289

87. Monograptus priodon - - 290

88. Various Graptolites - - 292

89. Zaphrentis konincki - - 299

90. Development of Major Septa in a Rugose Coral - 300

91. Septal Development in Rugosa - 301

92. Lithostrotion irregulare - 305

93. Parasmilia centralis - 307

94. Fossil Corals - - 312

95. Porifera - - • 317

96. Dimorphism in a Foraminifer - - 321

97. Nummulites - - 324

98. Orthophragmina umbilicata - - 325

99. Lepidocyclina tournoueri - 326
100. Fossil Plants - - - 336

PALEONTOLOGY

THE BRACHIOPODA

ON a large-scale geological map of England there is
seen a narrow, sinuous band of colour traceable almost
continuously from the Dorset coast to that of Yorkshire,
explained on the index as " Cornbrash." This corre-
sponds to the outcrop on the actual ground of beds
of rubbly limestone, in which are many small pits very
attractive to the fossil collector. Various kinds of fossils
may be picked up, but among the most noticeable, and
often the most beautifully preserved, are the forms repre-
sented in Figs, i and 2.

It is seen at once that these two forms, though
differing in details, have many features in common which
distinguish them from their companions in the Corn-
brash. Each one shows perfect symmetry about a
single plane ; each has at one end a rounded projection
perforated by a circular opening ; and the surface of
each, examined under a lens, shows, if in a good state of
preservation, a regular pattern of minute dots (Fig. i, c).

Although, as a rule, each specimen appears at first
sight to be a single solid body, examination soon shows

2 PAL/EONTOLOGY

that it is r3j.lly a ho!lo-.v shell (the test) composed of two
main portions in contact along their edges. These are
called the valves of the shell, and the valve which carries
the circular opening is called (for reasons to be seen
presently) either the ventral valve or the pedicle valve ; the

FIG. i.— TEREBRATULA INTERMEDIA (J. SOWERBY).

a, Dorsal view ; b, ventral view ; c, punctation ; dt side view ; e, umbo,
foramen, and deltidial plates of ventral valve, a, i>, d, Natural size ;
e, X2; c, greatly enlarged. (Original.)

other is called the dorsal or brdchial valve. The rounded
protuberance of the ventral valve is called its beak or
umbo. The dorsal valve has a similar, but much less
prominent umbo, without any opening. The end of the
shell at which lie these umbones is regarded as the hind
or posterior end, the opposite end being the front or

THE BRACHIOrODA 3

anterior ; the length of the shell is from one to the other.
The breadth of the shell is measured at right angles to
the length, in the plane separating the two valves, while
the thickness is from the surface of one valve to that
of the other where they are farthest apart.

FIG. 2. — ORNITHELLA OBOVATA (J. SOWERBY).

a, Dorsal view ; b, anterior view ; c, side view, all natural size ; d, umbonal
region of both valves, X f, showing cardinal area, foramen, and
deltidial plates in ventral valve, umbo and dark streak indicating
internal septum in dorsal valve. (Original.)

The surface is nearly smooth, but at irregular intervals
it is marked by fine lines nearly parallel to the valve
margins, but becoming crowded together as they approach
the umbo. Evidently these lines, like the rings of a
tree or those of a ram's horn, mark stages of growth
(growth-lines) ; they must, in lact, represent the margins

4 PALAEONTOLOGY

of the valves at successive intervals, and it is evident
that the earliest formed shell must be that in the region
of the umbo.

The region of the umbones is usually obscured by
rock substance, which it is a tedious matter to remove,
but when this is done some interesting features are
revealed. The anterior side of the circular opening
(foramen} is seen to be differently bounded from the
remaining three-quarters of its circumference. It is not
bounded by the ventral valve proper, but by the oblique
ends of two small plates of triangular shape : these are
called the deltidial plates, and the complete gap in the
shell (of which the foramen is only a part) which would
be exposed by their removal is called the dflthyrium.
On either side of these plates the surface of the ventral
valve shows a somewhat crescent-shaped, concave area,
having the appearance of being cut out from the shell :
this is the cardinal area. Along this area the margin of
the ventral valve is seen to be pressed close against and
slightly to overlap that of the dorsal valve : it is easy to
see that here the valves are hinged together, while else-
where they may have been free to gape apart, the gaping
being greatest at the anterior end. Indeed, although the
great majority of these fossils are found with the valves
tightly closed, one is occasionally found in which they
yawn more or less.

All the description so far given will apply equally well
not only to the two forms from the Cornbrash, but to
many others found in older or younger strata. All these
were long united under the common name Tevebvatula,

THE BRACHIOPODA 5

and were said to constitute a genus of animals. At the
present time the name Terebratula is used in a more
restricted sense, but we can still speak of all these forms
as terebratuloids. To distinguish these two species from
one another James Sowerby, nearly a century ago, named
one Terebratula intermedia and the other Terebratula obovata.
Let us now see some of the differences that underlie
their resemblances.

The two Cornbrash species obviously differ in pro-
portions. The ratio of length, breadth, thickness is in
T. intermedia roughly 12, 10, 6, in T. obovata n, 10, 8,
the latter being thus broader and thicker. In the
latter also the boundary between the cardinal area and
the general surface of the ventral valve is a very definite
edge, while in the former it is rounded off so as to be
somewhat indefinite. The foramen of T. intermedia is
much larger in proportion to the size of the whole shell,
and the outer edges of the deltidial plates, if continued,
would form an obtuse angle ; while in T. obovata the
foramen is small, and the corresponding edges would
form a right or slightly acute angle.

If we break open these two shells we shall often find
them to be hollow, and projecting into the cavity from
near the posterior end there is what looks like a looped
ribbon covered with little crystals of calcite. In the case
of T. obovata this loop extends forward nearly to the front
end of the cavity ; in T. intermedia it is less than half this
length (Fig. 3). In the former also there runs along the
middle line of the dorsal valve an internal ridge, forming
an incomplete partition (septum). The presence of this can

6 PALEONTOLOGY

be detected from the exterior by the appearance of a dark
streak : little or no trace of such can be seen externally
in T. intermedia. It has been found that the presence of
such a dark streak, indicating a strong dorsal septum, is
constantly associated with the presence of a long internal
loop, and its absence with a short loop, in all terebratu-

— So.

FIG. 3. — BRACHIAL LOOPS.

a, T. intermedia; b, b' , O. obovata (natural size); a, b, interior views of
dorsal valve ; b' , side view, test removed to expose the septum ;
H.P., hinge-plate; S., septum; So., sockets. (After Davidson.)

loids. We may therefore divide these into a long-looped
and a short-looped group.

This difference in the length of loop is regarded by
palaeontologists, for reasons that will appear later, as a
more fundamental difference than any of the others that
have been given. It is considered too important to be
expressed by a mere difference of specific name : the
generic (and even the family) names must also be

THE BRACHIOPO13A ?

different. Thus it is that the species obovata is no longer
called Terebratula, but Ornithella.*

Shells such as these are not only found fossil in many
different strata; they are also dredged, with the animal
still living of which they form part, from the ocean in
most parts of the world. Altogether 112 species of
terebratuloids have been found in the exploration of the
sea, as well as forty-six other species which, though not
actual terebratuloids, are so near them in structure that
they are united with them in zoological classification
under the name Brachiopoda (or bvachiopods). They are
purely marine in habitat, no species having ever been
found in any lake, river, or estuary ; the largest number
of species live on the bottom between the depths of 30
and 100 fathoms ; many species (and perhaps the greatest
number of individuals) live in shallower water than this,
even in a few cases above low- water mark ; only a few
species live in water beyond the loo-fathom line. The
coasts of Japan furnish the largest number of species for
any one area : those on the east and west coasts being
almost entirely different (20 species in eastern, n in
western waters, of which only 2 species are common
to both.) The geographical range of some species is
very wide, that of others very restricted.

A living terebratuloid, such as Magellania flavescens of
the Southern Seas, lives attached to the sea bottom
(usually many individuals in a cluster) by a short
muscular stalk (pedicle) which emerges through the

* For further explanation of the terms species, genus, family, etc.,
see Chap. XII.

8 PAL/EONTOLOGV

foramen in the ventral valve. It feeds entirely upon
the microscopic organisms, both animals and plants (but
probably in the main the minute plants known as
diatoms) which drift about helpless in the currents in
the water. So enormously abundant are such organisms
in the ocean, and so quickly do they grow and reproduce
themselves, that they form the food, at first, second, or
third hand, of nearly all the animals in the sea. It has
well been said that " all fish is diatom " in the sense that
" all flesh is grass," for there is nothing in the ocean
comparable to the vegetation of the land. The larger
sea-weeds do not afford food to many of the marine
animals.

An animal feeding upon such helpless microscopic
prey can do without many of the organs that are essential
to the familiar animals of the dry land — eyes and ears to
see their prey, arms and jaws to seize them, swimming or
running organs to pursue them, are all needless. Shut
up within its shell a brachiopod is comparatively safe
from enemies ; fixed by its pedicle, only very violent
storms can throw it up to die on the shore. All it needs
in external organs are (i) some means of creating a
constant current into and out of its shell-cavity, to bring
in food and oxygen and carry out waste products ; (2) a
shell so arranged that it can be closed tightly when
necessary for safety, and opened at will to allow ingress
and egress of the water-currents. These necessities
control and determine the structure both of the soft parts
of the animal's body and of the shell which it secretes.
The actual " body " of the terebratuloid, containing

THE BRACHIOPODA 9

the digestive tube, central nervous system and other
fundamental organs, occupies a surprisingly small portion
of the interior of the shell. From it there arise a pair of
muscular sheets which form a lining to the two valves,
and by which the valves are in fact secreted : these are
called the mantle-lobes, and the cavity enclosed by them
is called the mantle-chamber. Hanging freely in this
mantle-chamber are a pair of spiral ribbons, fringed with
tentacles along their whole length and covered with
microscopic vibratile hairs (cilia), the constant movement
of which creates the water-current which is so vitally
necessary, and also serve as gills (breathing organs for
the exchange of gases between blood and sea-water).
The shelly " loop " which we have seen in the fossil
forms serves to support these spiral " arms " (or brachia)
as they are misleadingly termed. In M. flavescens the
loop is long and doubled back on itself, the two halves
of this doubled-back portion being distinguished as
"ascending lamellae" from the longer "descending
lamellae" (Fig. 8).

If we examine the interior of the valves of a modern
terebratuloid, we can make out certain details of structure
more easily than on a fossil specimen; though after-
wards we shall find we can recognize the same details in
the fossils. The two valves are hinged together by
means of a pair of projections from the ventral valve
(hinge-teeth) which fit into two hollows (dental sockets) in a
horizontal shelf in front of the umbo of the dorsal valve
(hinge-plate). To give leverage for the opening of the
valves there is a short knob-like projection (cardinal

io PALEONTOLOGY

process) from the dorsal valve into the umbonal cavity of
the ventral valve. The line along which the margins of
the two valves meet in the neighbourhood of the hinge-
teeth is called the hinge-line. To effect the opening of
the valves there are muscles fixed at one end to the
cardinal process, at the other to the interior of the
ventral valve (divaricatov muscles) ; to close the valves
there are other muscles running across from valve to
valve (adductor muscles). All muscles act by contraction
which pulls their two attachments towards one another.
The dorsal attachments of these two sets of muscles
being on opposite sides of the fulcrum (the hinge-line),
they work against one another. The muscles of the
pedicle are attached to the interior of the ventral valve.

When the muscles decay after death, the areas of the
shell to which they were attached show as slight depres-
sions devoid of the smoothness of the rest of the internal
surface. Thus it is possible to tell in a fossil brachiopod
the presence and arrangement of these muscles. Although
the details differ greatly among the many forms of
brachiopods, the adductor impressions in the ventral
valve are always close together, and the divaricator im-
pressions lie more or less to the right and left of them.

The body proper of a brachiopod being so small, it is
not surprising that the reproductive organs should extend
into the mantle which lines the two valves ; impressions
of the ovaries are in many forms to be seen on the inner
surface of the shell ; so also are the branching impres-
sions of vessels belonging to a circulatory system
(vascular markings). In forms like the terebratuloids

THE BRACHIOPODA 11

which have a punctate shell, each punctation contains a
minute outgrowth of the mantle, and in at least one
recent species this has been shown to have the structure
of a simple sense-organ, though the nature of the sense
is unknown. m

From the egg the larval brachiopod escapes as a
microscopic body covered with vibratile cilia, by means
of which and the action of currents it gets the only
chance of its life to migrate away from the fixed home of
its parents. It is not likely that any one generation
travels very far, yet some modern species are world- wide,
thanks to the cumulative effect of small journeys through
thousands of generations. The length of life of brachio-
pods in general is not known, but in one species there is
evidence of five years' life.

Very soon the larva fixes itself by a rudimentary
pedicle and begins to secrete two valves. In all observed
cases this first shell, or protegulum, has the same general
form, which is also the form of the most primitive
brachiopod shells in the Lower Cambrian rocks (Fig. 6, a).
This resemblance favours the view that the development
of the individual (ontogeny) repeats, in an abbreviated
and more or less imperfect manner, the ancestral history
of the race (phytogeny). This principle of recapitulation, or
palingenesis, is of the utmost importance in palaeontology
as a criterion of blood-relationship, and especially in the
case of those animals whose shell-growth is of the kind
found in brachiopods, where the early stages of the shell
are preserved and added to, to form the later (growth by
accretion).

12 PALAEONTOLOGY

The protegulum of a brachiopod forms the apex of
each valve. Too often, owing to the exposed position of
the iimbo, especially of the ventral valve, it may be
rubbed away in later life; in some forms it is absorbed
in the enlargement of the pedicle-opening ; but in many
cases it persists through life. Growth consists in adding
to it both on its inner surface and on its margins — to
the greatest extent at the anterior margins. During this
growth the shape of the shell may change, perhaps more
than once ; it may acquire surface-ornament, and this
may alter or disappear ; but (apart from wear and tear)
every stage is preserved in the adult shell, and if the
characters of an early stage are the same as those of the
adult shell of a brachiopod of earlier date, there is a
presumption that the latter may be ancestral to the one
under consideration.

Growth by accretion is not possible, however, for all
parts of the skeleton : the loop, once formed, could not
grow so and remain a loop. As the valves increase in
size, repeated resorption and fresh secretion of the shelly
substance of the loop must take place. The difference
in development between the valves and the loop is like
the difference between that of a ram's horn and that of a
stag's antler. Hence the adult loop affords no evidence
of the stages through which it has passed. These can
only be determined by actual observation of the develop-
ment in living brachiopods, and by the fortunate preser-
vation of immature as well as mature specimens among
fossils. The former method has yielded results which
afford striking confirmation of the principle of palin-

THE BRACHIOPODA 13

genesis, for it has been shown by Beech'er and Thiele
that the loops of the modern terebratuloids Magellania and
Macandrevia, though much alike in the adult stage, pass
through two quite different series of metamorphoses, and
in each case the successive stages correspond to the
adult loop in a number of other genera, some still living,
others extinct (Fig. 9).

It will be convenient here, before returning to the

6.

FIG. 4.— HEMITHYRIS PSITTACEA (CHEMNITZ).
Recent, x-|.

a, Ventral ; b, dorsal valve; Add., adductor muscle-impressions ; C, crus ;
Del., delthyrium, the deltidial plates removed; Div., divaricator
impressions; D.P., dental plates; S, socket; T, hinge -tooth.
(Original.)

fossil forms, to consider some of the other brachiopods
found living in the sea.

Hemithyris psittacea (Fig. 4) is a form living in the cold
northern waters : its shells are often cast up by storms on
the coast of Labrador. It agrees with the terebratuloids
in having dissimilar symmetrical valves, the larger being
perforated for a pedicle, and in having the delthyrium
partly closed in by a pair of deltidial plates. But these
plates, instead of bounding the foramen on its anterior
side only, bound it laterally and even tend to meet on
the posterior side just under the umbo (as in Fig. 9). On

14 PAL/EONTOLOGV

examining the surface with a lens we fail to find any trace
of the pattern of punctations so characteristic of the
terebratuloids; but we observe a silky, fibrous appear-
ance. Internally, we find still more striking differences.
It is true that the cardinal process, hinge-plate and
sockets, hinge-teeth and the several pairs of muscle-
impressions are all present with only slight differences of
arrangement ; but although spiral brachia exist there is
no calcareous loop. From the hinge-plate two short
processes project into the cavity, and correspond to the
beginning of a loop : they are termed the crura (plural
of cms). Again, in the umbonal cavity of the ventral
valve there are a pair of vertical partitions which, as
they extend into and strengthen the teeth, are called
dental plates : these, however, are also found in some
terebratuloids.

Forms more or less similar to Hemithyvis psittacea are
often found fossil, in the Cornbrash among other strata.
In the works of Sowerby they were included under the
broad name of Terebratula, but afterwards distinguished
by the separate name Rhynchonella by Fischer von
Waldheim. This name in turn has had to be restricted
to a portion only of these forms, but we can still con-
veniently speak of them as rhynchonellids .

Thecidea (Lacazella) meditevvanea is a small brachiopod
with a very thick shell. It has no pedicle, but is found
cemented to other shells by its ventral umbo. The long
and narrow delthyrium occupies the middle of a large
cardinal area, and is completely closed up, not by a pair
of plates, but by a single one (deltidium). Internally there

THE BRACHIOPODA 15

are no crura even, but the thick dorsal valve has deep
furrows with ridges between corresponding to the course
of the spiral brachia. The other structures do not differ
essentially from those of rhynchonellids and terebratuloids,
but evidently. this is something distinct from both. In
particular, the single deltidium in place of two deltidial
plates is a more important difference than might at first
sight appear, because the study of the early development
has shown that these structures appear as secretions by
different parts of the body — the single plate, by the pedicle
itself; the double plates, by the ventral mantle-lobe.
Only two species of Thecidea are known to be living now,
but many fossil species are known, though on account of
their small size they are often overlooked. Among modern
forms Thecidea is an isolated genus with no near allies ;
but there are many extinct forms more nearly related to
it than to any other living genus.

Discinisca lamellosa (Fig. 7, c) is found in shallow water
or between tide-marks on the coast of Peru. Its shell
differs both in texture and in shape from those hitherto
considered, for it is not calcareous but has a horny
appearance and consists of organic matter with some
calcium phosphate instead of carbonate ; and each valve
is nearly circular in outline, with the umbo near the
centre. The ventral valve resembles a low cone of
which one portion of the surface (on the posterior side)
has been crushed in. In the middle of this depressed
area is a narrow slit through which the pedicle passes
obliquely. The dorsal valve is a still flatter cone.
Internally, there are no hinge-teeth, no cardinal process

1 6 PAL/KONTOLOGY

or hinge-plates, no crura or loop ; in fact, the only
internal features of the shell are the muscle-impres-
sions and a small median septum in the ventral valve
anterior to the delthyrium. Here is a brachiopod differ-
ing greatly from all the previous forms, especially in not
having its valves hinged together, but only attached by
muscles and other soft tissues. There are six species
of Discinisca living and one species of the allied genus
Discina : they may be spoken of as discinids. Many
fossil discinids are known, but owing to the inarticulate
(unhinged) character of the shell it is commoner to find
isolated valves than complete shells — the reverse is the
case in terebratuloids and rhynchonellids, and indeed
most brachiopods.

Lingula anatina (Fig. 7, b) is a brachiopod which
burrows in the sea-bottom in very shallow water; it
has a very long pedicle which it uses to draw itself
into its burrow when necessary. Like Discinisca it has
a horny shell. In shape it is somewhat oblong, the two
valves being very nearly alike, each having the umbo at
the extreme posterior end. There is no delthyrium, the
valves diverging at the posterior end for the passage
of the pedicle. Internally the valves show no hinge-
structures or brachial skeleton, though the spiral brachia
are present as usual ; but in the dorsal valve there is just
below the umbo a reflected portion of the shell resembling
a cardinal area. The muscular impressions are faint,
but are more complex than in previous cases, the absence
of articulation making it possible to have muscles for
sliding the valves sideways as well as for opening and

THE BRACHIOPODA 17

closing the shell. Fifteen species of living lingulids are
known, and a very great number of fossil species.

The resemblance of many fossil brachiopod shells to
those of living forms compels us to believe that the
former were also once parts of living organisms which
lived in much the same manner. Where therefore we
find brachiopod shells in a stratum we must conclude
that it was originally a deposit on the sea-bottom, and
we may even venture to estimate the depth below the
surface at which it was deposited from a consideration of
the abundance and generic identity of the brachiopods.
But the fossils also serve another very important
purpose.

As long ago as 1688, the famous Dr. Robert Hooke
foresaw the possible utility of fossils as time-markers.
" However trivial a thing a rotten shell may appear to
some," he wrote, " yet these monuments of nature are
more certain tokens of antiquity than coins or medals . . .
and though it must be granted that it is very difficult to
read them and to raise a chronology out of them . . .
yet it is not impossible." It was William Smith who, at
the end of the eighteenth century, was first able to " raise
a chronology out of them," by showing that in the series
of strata that lie one upon the other in the Bath district
(and elsewhere) each division is distinguished by particular
species of fossils.

Thus William Smith found that underneath the Corn-
brash were a series of beds of hard flaggy limestone,
which he termed the " Forest Marble " ; below these again

i8 PALEONTOLOGY

the Bradford clay ; below this the thick mass of oolitic
limestone for which Bath is famous, and so on. In
Fig. 5 are shown the species of brachiopods that are
characteristic of two of these formations. The value of
such fossils for arranging the stratified rocks of the earth
in systematic order will be realized when we remember
that though in a hilly district with many quarries and
other openings into the rocks like that around Bath, or
along a cliffy coast, it may be quite easy to prove clearly
how one group of strata overlies another, yet when we
try to follow the strata from place to place and especially
into regions where exposures of the rocks are few and
poor, there are many difficulties to be met. A stratum
may change its character, for example the Forest Marble
changes from limestone to clay in Northamptonshire ;
or it may "thin out" and disappear altogether; or a
"fault" may suddenly shift its position greatly. Amid
these difficulties, the finding of distinctive fossils will
often save the geological surveyor from a mistake which
might perhaps lead someone to sink a mine or bore for
water in a wrong place.

The whole group of animal species found together in
the same beds are spoken of as the fauna of those beds,
just as the collection of species now living in any given
area is the fauna of that area. In past times, as at
present, a fauna had a definite habitat or distribution in
space ; so that the fact of a fossil fauna in Australia being
different from one in England does not prove that they
were not contemporaneous. This fact may at first sight
seem to make it impossible to apply William Smith's

THE BRACHIOPODA 19

principle over great distances; but how it can be done
will be explained later. At present we must consider
another point.

At least 6,000 species of fossil brachiopods have been
described, and of these only 25 are identical with
living species, while 133 living forms are unknown as
fossils. These 6,000 species represent the members of a
great number of successive faunas that have succeeded
one another throughout geological time. If species
followed one another without any guiding principle of
succession, the human memory would be unable to
grapple with the geological sequence. Fortunately this
is not the case. While species are short-lived, geologic-
ally speaking, genera are longer lived, but yet most of
them are restricted to a limited portion of geological
time ; and the abundance of certain genera will char-
acterize large groups of strata, just as the abundance of
certain species marks a smaller group. The smallest
thickness of rock that can be distinguished by means of
its fossils over a wide geographical area is called a zone :
its actual thickness may vary in different cases from an
inch or even less to several hundred feet, according to
the net rate at which sediment was accumulating. A
zone can, if not very thin, be divided into a number of
beds. The time during which the strata constituting a
zone were deposited is called a hemera, and is denned as
the time during which some one selected species was at
the acme or height of its development. We are unable
at present to bring this unit of geological time into
relation with any of our astronomical time divisions.

20 PALEONTOLOGY

A number of consecutive hemeva (varying in different
cases from three to ten) are grouped as an age, correspond-
ing to the development of a genus or a family : the
corresponding thickness of strata is called a stage. Ages
are again grouped into epochs, these into periods, and these
into eras, the bounds of these being determined by great
and rapid changes of fauna. These terms and their
equivalents may be tabulated thus :

TiME-DlVISIONS. ROCK-DIVISIONS.

Era. Group.

Period. System.

Epoch. Series.

Age. Stage.

Hernera. Zone.

Thus the Cornbrash is, approximately, the zone of
Terebratula intermedia. Each of the brachiopods shown in
Fig. 5 also gives its name to a zone, and all these zones
with some others make up the Bathonian stage, which
takes its name from Bath, around which city these zones
are well displayed. This stage is part of the Upper
Jurassic Series of the Jurassic System, so named from
the Jura Mountains. This in turn is part of the Mesozoic
Group.*

The sequence of forms in division after division of
geological time often appears casual and meaningless,
but that is probably only owing to the imperfection of
our knowledge. In a number of cases we can trace a

* See Appendix I. for tabulated statements of time-divisions.
Reference to these tables will frequently be necessary during the
reading of what follows.

THE BRACHIOPODA 21

succession which has all the appearance of being an
actual genealogical tree, and as we pass back into the
earlier geological times we find a convergence of trie
lines of descent towards a few very simple forms. At
the beginning of the palaeontological record we seem to
be very near the beginning of the Brachiopoda.

Classification must be based on structure and blood-
relationship. We cannot at this stage give a full justifi-
cation for the following classification (which is essentially
that of Beecher, 1893), but it is given as the best of

^ \

FIG. 5. — BATHONIAN TEREBRATULACEA.

a, a', Epithyris bathonica S. Buckman, Bath oolite, x|; b, Ornithella
digona (J. Sowerby), Bradford clay, xf. (After Davidson.)

many attempts to express the inter-relationships of
Brachiopoda.

The Brachiopoda are so well defined and sharply
marked off from all other animals that they may well be
accorded the rank of a phylum or primary branch of the
Animal Kingdom.* Brachiopoda are marine animals
fixed by a pedicle or otherwise, feeding on microscopic
floating organisms by means of spiral " brachia,"

* Many zoologists unite the Brachiopoda with the Bryozoa in a
phylum Molluscoidea, of which they form two classes. The differ-
ences between them appear to the present writer too profound to be
expressed as merely class differences.

22 PALEONTOLOGY

secreting a shell with dorsal and ventral dissimilar
symmetrical valves.

In the opinion of most students of the phylum, the
most fundamental plane of cleavage is that between
forms like Lingula and Discinisca, in which the valves are
not articulated by teeth and sockets, and those like
Ttrebratula in which they are. These two divisions con-
stitute the class Inarticulata and the class Articulata
respectively.* The characters of the delthyrium afford
the next means of subdivision. Inarticulata such as
Lingula which have no delthyrium constitute the order
Atremata ; those with a delthyrium enclosed by the
ventral valve (at least during early life), the Neotvemata ;
Articulata with a median deltidium form the order
Protremata ; those with a pair of deltidial plates, Telo-
tremata. Each of these orders is divided into silb-
orders (or, as Prof. Schuchert prefers to call them, super-
families) and these into families, made up in turn of
one or more genera, each with one or more species.

The beginner in Palaeontology is often troubled by the
importance attached in classification to the internal
characters of fossils or other features which are often
invisible in ordinary specimens, and is puzzled to under-
stand how identification is possible in such cases. It
may be pointed out that if one single specimen shows all
the characters that are necessary to determine its
systematic position, many others can be recognized as

* Professor Schuchert, of Yale University, dissents from this
view, regarding articulation as a feature that has been evolved
independently in the orders Protremata and Telotremata.

THE BRACHIOPODA 23

identical with it by their minor features. Hence the
apparent paradox that it is often easier to determine the
species of a fossil than its genus, which is quite contrary
to the ordinary conception of classification. For the
specific character is commonly some feature of external
shape or ornament that is easily recognized but might be
found in any one of a number of genera, while the generic
character may be some feature which only an excep-
tionally well-preserved specimen will show. We must
never forget that classifications are not framed as means
for the easy naming of specimens, but are intended to
indicate the real natural relationships of species.

In the particular case of the internal characters of
brachiopods, however, these are not so inaccessible as
a beginner may think. It often happens that the interior
of a shell may be filled, after death, with foreign material
such as mud or sand, and that afterwards the shell may
be dissolved, leaving the consolidated mud or sand as in
internal cast or mould of the shell. On such a cast all
the structures on the inner surface can be recognized —
hollows being represented by elevations, and projections
by depressions. It is even advisable, when the infilling
material is suitable and specimens are abundant, to make
casts artificially by burning away or otherwise removing
the shell.

ORDER I. : ATREMATA.

Inarticulate Brachiopoda with horny and phosphatic
shells, the pedicle emerging between the divergent valves,
which show little or no tendency to enclose it in a
deltriyriutru

24 PALEONTOLOGY

This is the most primitive order, especially character-
istic of the Older Palaeozoic era, only a single family, the
Lingulida, surviving through later eras down to the
present day. It is divided into two sub-orders :

1. Obolacea. — Rounded and lenticular in form, with
short pedicle and thick shell ; probably fixed to floating
seaweed. These are confined to the Older Palaeozoic
era, and range from the most primitive of all brachio-
pods, Rnstella of the Cambrian, which may be near the
ancestral form of all four orders, through the Obolidce. to
the highly specialized Trimerettida of the Ordovician
and Silurian. In these last, the muscle attachments are
carried on platforms separated from the inner surface of
the valves by long and narrow cavities : these give the
internal casts a very complex form. In Obolus, each
valve has a well-marked cardinal area, that of the ventral
valve with a longitudinal pedicle-groove, and its umbo
being more prominent.

2. Lingulacea. — Elongated in form, burrowing, with
thin shell and long pedicle.

These begin in the Cambrian with Linguklla, which
retains -the obolid characters of a cardinal area, with
pedicle-groove in the dorsal valve only, and of a more
prominent ventral umbo. It is followed in the Ordovician
by Lingula, in which the valves are almost equal, and the
pedicle groove is wanting. This genus survives to the
present time, and so is one of the longest-lived of all
genera, not only of Brachiopoda, but of all but the
lowliest phylum of animals.

ORDER II.: NEOTREMATA.

Inarticulate Brachiopoda, in which the valves acquire
a more conical form, and the pedicle becomes surrounded
by the ventral valve and comes to emerge close to the

THE BRACHIOPODA 25

umbo and at a distance from the margin. In the more
typical sub-orders the shell is horny and phosphatic.

SUB-ORDER i. Acrotretacea. — Shells but little
changed from obolids in shape, especially the dorsal
valve. Older Palaeozoic. Chief genera : Acrotreta, in
which the ventral valve has a high cardinal area (Camb.-
Ord.),* and Siphonotreta in which the pedicle-opening
leads into an internal tube, and the shell is perforated by
pores which extend into hollow spines (Ord.-Sil.).

SUB-ORDER 2. Discinacea. — With each valve more or
less conical in shape, umbo eccentric. In Trematis of the
Ordovician the pedicle-aperture is a long slit extending
from the posterior margin to the umbo ; but in Orbicu-
loidea (Ord.-Cret.) it becomes closed-in as growth pro-
ceeds by a plate called the listrium, and the short pedicle
passes obliquely through the slit that remains. In exist-
ing seas this sub-order is represented by Discinisca and
Discina.

SUB-ORDER 3. Craniacea. — Aberrant forms, in which
the shell is calcareous and the pedicle and pedicle-
opening are lost, fixation by cementation taking the
place of fixation by pedicle. Ordovician to Recent. The
principal genus Crania has the range of the sub-order.

ORDER III. : PROTREMATA.

Articulate Brachiopoda, with delthyrium restricted by
a deltidium. Shells calcareous. No brachial skeleton,
except crura in some cases.

SUB-ORDER i. Strophomenacea.— In the Cambrian,
this is represented by the little Kutorgina, which, but for
its more calcareous shell, broader shape and wide delthy-
rium, differs little from the contemporary Inarticulata, for

" These are obvious abbreviations of the names of the geological
systems, tabulated in Appendix I.

26

PALEONTOLOGY

its deltidium and hinge-teeth are rudimentary. It shows,
however, the long straight hinge-line and wide shell that
characterize so many of the Palaeozoic Articulata.

In the Ordovician, and still more in the Silurian, this
sub-order attains very great importance, and some of its
most familiar members deserve detailed treatment.

Leptana rhowboidalis (Fig. 6) of the Wenlock Limestone

V.A.

C.P.

-D.A.

FIG. 6. — LEPT^NA RHOMBOIDALIS, WILCKENS.
(Natural size.)

a, Dorsal view ; b, interior of ventral valve ; r, posterior view ; d, interior
of dorsal valve; Add., adductor impression ; Ch., chilidium (showing
median pedicle-groove); C.P., cardinal process; D, delihyrium ;
D.A., cardinal area of dorsal valve ; Div., divaricator impression;
V.A., cardinal area of ventral valve; V. U.t umbo of vtntral valve.
(Original.)

is a broad shell (length : breadth = about 2:3) with a semi-
elliptical outline, the long straight hinge-line forming the
greatest breadth of the shell. The early (posterior) part
of the shell is flat, but after growing to a length of about
two centimetres a sudden change takes place and the

THE BRACHIOPODA 27

surface of the ventral valve becomes bent on itself at
about right angles, making that valve in the adult highly
convex externally, while the ventral valve is correspond-
ingly concave, so that the change does not result in great
increase of thickness.* Concavo-convex shells are
common among Palaeozoic brachiopods : when, as in
Leptana, the dorsal valve is concave, they are said to be
normally concavo-convex ; when the ventral is concave they
are reversed concavo-convex (or convexi-concave).

The surface of the shell is ornamented with radiating
ridges, crossed by much finer concentric lines. In addition
to these there appear at a little distance from the umbo
coarser and irregular concentric corrugations, which
rapidly increase in size up to the' line of growth-change
(reflection), beyond which they cease, as though they
were premature attempts to make the change in growth -
direction.

On each valve there is a wide and low cardinal area,
extending all along the hinge-line. The delthyrium
notches both areas equally so that it is rhomboidal in
shape. In the ventral valve there is a small deltidium
close to the umbo ; in the dorsal a much larger convex
plate (chilidium\ with a median groove, covers the whole
opening. The pedicle-opening is small, between umbo
and deltidium (encroaching on the former in old age) ;
the pedicle probably rested on the groove of the chilidium.
In the interior of the ventral valve, well-marked muscle-

* In Ornithella obovata a somewhat similar growth-change takes
place in old age, but that being a biconvex shell, both valves increase
in external convexity, and so the shell becomes greatly thickened.

28 PALAEONTOLOGY

impressions and teeth are seen. The raised rims of the
divaricator-areas join on to the teeth and delthyrial
margin like rudimentary dental plates, and there is a
slight median septum. In the dorsal valve, the cardinal
process is short and bifid and fits closely under the
chilidium. Dental sockets are conspicuous.

Allied genera to Leptcena are Rafinesquina (with fine
radial striations alternately larger and smaller, and with
ridged muscle-areas) and Stvopheodonta (with fine denticu-
lations along the hinge-line, and inconspicuous chilidium).
Neither of these has the concentric corrugations of
Leptana. All three are represented by common species
in the Ordovician and Silurian systems. Davidsonia of
the Devonian fixed itself by cementation on to other
shells, and shows internal spiral markings which are
interesting as evidence that spiral brachia existed in this
extinct family.

Of genera with reversed concavity, Strophomena might
be described as a reversed Rafinesquina^ and StropJwnella
as a reversed Stvopheodonta.

The sub-family Orthotetina (Fig. 7, f-i) is characterized?
in general, by well-marked dental plates, which may
extend far forward in the pedicle-valve — sometimes
parallel to one another (Meekelia), or diverging (Schell-
wienella) or converging to form a median septum
(Ovthotetes). In Derbya, however, the dental plates are
greatly reduced, the septum extending to the apex of
the umbo. These four genera are Carboniferous and
Permian. Externally all the genera of this sub-family
are radially ribbed. Certain species belonging to different
genera are so closely similar externally, that it is im-
possible to discriminate between them without examining
the interior, for which the grinding of sections may be
necessary : such species are said to be homoeomorphs of

THE BRACHIOPODA 29

one another. The failure to discriminate between
homceomorphs has frequently led to mistakes in the
correlation of strata.

Productus productus (commonly, but wrongly, termed
Productus martini) is one of many species of Pvoductus that
abound in the Lower Carboniferous Limestones and
Shales. In general form it is not unlike Leptcena, but the
convex ventral valve shows a stronger and more uniform
curvature, and its umbonal region is much larger and
more rounded ; there is no cardinal area and no delthy-
rium. The dorsal valve is for the first two centimetres
from the umbo, almost flat, very slightly concave ; it also
has no cardinal area and its umbo does not project above
the hinge-line. The hinge-line is long and straight, but
less than the greatest breadth of the shell. The profile
of the ventral valve first forms an arc which is almost a
semicircle, of which the dorsal valve forms the chord, and
at two centimetres from the umbo the valves meet as
though the shell were complete. From this distance on,
however, the ventral valve is continued with very slight
curvature for a considerable length, while the dorsal valve
bends abruptly and continues in close contact with the
ventral, the two valves thus forming a sort of flange with
no space between them when the shell is closed. This
flange is very easily broken off during extraction of the
fossil from the rock, and what remains has all the
appearance of a perfect shell, two centimetres long and
about 2*5 centimetres broad.

The surface is ornamented with close-set radiating
ribs, rounded in section, with a few concentric corruga-

PALEONTOLOGY

FIG. 7 — VARIOUS BRACHIOPODS.

a, a', Obolus appolinis Eichwald, Cambrian, ventral, and dorsal valves :
below the umbo in each is the cardinal area (shaded), that of the
ventral valve having a median pedicle-groove ; muscle-impressions in
outline. (Natural size.) b, b' ', Lingula anatina Bruguiere, Recent,
ventral, and dorsal valves: parts as in a, a'. (x^t.) c, c', Dis-
cinisca lamcllosa Broderip, Recent, ventral, and side views. Fora-
men black. (Natural size.) d, d' , Dalmanella [Orthis~] elegantula
(Dalman), Silurian, ventral, and dorsal valves: muscle-impressions
shaded, (xf.) e, e't Productus semireticulatus (Martin), Lower
Carboniferous, dorsal view, and interior of dorsal valve, (x^.)
/, g, Umbonal views of ventral valves of f, Schellwienella ; g, Derbya;
areas shaded, dental plates (black) showng through the test. A, Urn-

THE BRACHIOPODA 31

tions near the umbones. Here and there, especially near
the sides, may be seen round spots, proved by comparison
with better-preserved specimens to be the stumps of
long curved spines. These spines are the special character-
istics of the family Productidtz, and they served perhaps to
anchor these brachiopods to the stems of crinoids or other
suitable objects, or perhaps to support them like springs
on the sea-bottom, thus enabling them to dispense with a
pedicle.

Internally, Productus does not differ greatly from
Leptana : the divaricator-impressions in the ventral valve
are more widely separated, and in front of each one there
is a spiral thickening of the interior of the valve like that
of Davidsonia ; in the dorsal valve there is a pair of some-
what similar, reniform (kidney-shaped) elevations.

In the genus Chonetes (Ord.-Perm.) we seem to
see a link with the Leptcenida. It has similar spines to
Productus on the cardinal margin of its ventral valve only,
and it retains cardinal areas and delthyrium. On the
other hand, Etheridgina (known only in the Carboniferous
of Scotland) shows a further modification of the Productus
method of fixation : the spines grip the crinoid stem
tightly, and the ventral valve becomes cemented to it.

FIG. 7. — VARIOUS BRACHIOPODS (continued}

bonal region of Orthotetes, ground down to show dental plates and
septum (black), i, Internal cast of ventral valve of Meekella, dental
plates black, j, Conchidium knighti (]. Sowtrby), Silurian, dorsal
view. /, Naturally split section ; /', cross-section, (xj.) k, Spirifer
striatus (Martin), Lower Carboniferous, dorsal view, with test paitly
removed, showing spiralia. (xj.) /, Cyrtia exporrecta (Wahlen-
berg), Silurian, dorsal view, (xj.) m, m', Atrypa reticularis
(Linne"), Silurian, side view and cross-section showing spiralia. (x§.)
A, Cardinal area; C, crus ; C.P., cardinal process; del., deltidium ;
D.V., dorsal valve; H.P., crural plates (cruralium) ; ./?, reniform
impression ; S, septum ; SP, spondylium (dental plates) ; T, hinge-
tooth ; V.V., ventral valve; X, septal plate. <z, b, After Walcott ;
f-i, after Thomas ; the rest after Davidson.

32 PALEONTOLOGY

In the Permian Strophalosia also, the ventral valve is
cemented to foreign bodies.

The family Orthida may be illustrated by tfalmanella
elegantula of the Wenlock Limestone, a small shell, nearly
circular in outline, little over a centimetre long and about
as broad (Fig. 7, d). The length of the hinge-line is less
than the greatest breadth. The ventral valve is very con-
vex, with a well-defined slightly concave cardinal area, a
triangular delthyrium without any deltidium, and teeth
supported by dental plates. The dorsal valve is very
slightly convex, with a very narrow area, and a very small,
bifid cardinal process occupying the centre of a small
delthyrium and at the end of a slight median septum ; the
inner side of each dental socket is produced into what at
first sight looks like a tooth, but as these do not fit into
sockets in the ventral valve they cannot be teeth but must
be crura like those of the rhynchonellids. The surface is
marked by rather fine radial ribs, which here and there
increase in number by bifurcation.

The absence of deltidium and presence of crura form
two important differences from the Strophomenacea and
appear to justify raising the orthids to the rank of a sub-
order— Orthacea.

SUB -ORDER 2. Pentameracea. — The uppermost
Ordovician and lowest Silurian beds in parts of England
and Wales are much alike, both consisting of calcareous
sandstones often crowded with fossils which are largely in
the form of internal casts. Many of these fossils are com-
mon to the two systems, or differ by characters too slight
to be detected in the field ; but the geological surveyor

THE BRACHIOPODA 33

knows at once that he has Silurian and not Ordovician
rocks before him when he sees casts of rounded outline
cut almost in half by deep fissures. These fissures in
the cast correspond to large internal plates in the shell,
such as only occur in the pentamerids, a group of
brachiopods which suddenly migrated into the British
area at the opening of the Silurian period.

One of the most familiar pentamerids is Conchidium
[Pentamerus] knighti (Fig. 7, j) of the Aymestry Limestone.
This is a large shell, attaining a length of 7*5 c.m., and
then about 4 c.m. broad and 5'5 c.m. thick. Both valves
are highly convex, with the umbo of each greatly curved
over ; there is no cardinal area ; the delthyrium is broad
and bears a concave deltidium (often wanting) ; the hinge-
line is short and curved. The surface is marked by coarse
rounded ribs which increase in number sometimes by
bifurcation, sometimes by intercalation of a new rib.
There are no regular concentric markings, but occasional
stoppages in growth are indicated by growth-lines.

It is not easy to extract a perfect specimen, because
the fossil splits so readily along the large internal plates ;
and it is not difficult to get an almost median section. By
combining what we see in such a section with the evidence
of a transverse section (Fig. 7, /, /') we can realize
that the mantle-cavity in the umbonal region and for
some distance forwards was divided into three chambers,
the middle one containing the body proper, the lateral
chambers probably containing the spiral brachia. To-
wards the anterior end the three chambers united. In
the ventral valve there is first a large median plate, a

3

34 PALEONTOLOGY

great development of the little septum seen in Leptana.
At its inner edge this is continuous with a pair of diverg-
ing plates which on being traced to the hinge can be
recognized as greatly-developed dental plates. We have
already seen that in Leptcena the rims of the muscle-areas
are continuous with the dental plates, and it is evident
that in Conchidinm the development of these plates must
lift the muscles right off the inner surface of the shell
into the median chamber. Such a pair of muscle-bearing
dental plates is called a spondylium.

In the dorsal valve instead of a median septum there is
a pair of septal plates a little on either side of the middle
line; from these diverge at a low angle another pair of
cnival plates the free edges of which are in contact with
those of the dental plates for some distance from the
hinge, thus bounding the central chamber.

Although Conchidinm knighti is commonly referred to
the genus Pentamerus, that should only contain smooth
species (e.g , P. oblongus of the Upper Valentian or
Llandovery stage). Other genera closely allied are
Stvicklandinia (e.g., S. lens, of the Lower Valentian or
Llandovery), with straight hinge-line and without the
greatly-curved umbones, and Gypidula (e.g., G. galeata,
Wenlock Limestone) in which there is a median elevation
(fold) on the ventral valve, and corresponding depression
(sinus) on the dorsal, an arrangement found in many
brachiopods, and serving to separate the lateral areas of
the mantle, where the inhalent current enters, from the
central exhalent area. Less close is Camarofhoria
(Dev.-Perm.) with the closest external resemblance
to a rhynchonellid, from which it may readily be
distinguished by the internal plates (seen through the

THE BRACHIOPODA 35

translucent shell or, better, on an internal cast), and by
the absence of the small foramen and deltidial plates of
the rhynchonellids. In this genus there is a dorsal as
well as a ventral median septum.

The Pentameracea appear to spring from the Cambrian
Syntrophia. Other primitive genera (Clitambonites, Poram-
bonites) occur in the Ordovician of Russia ; the typical
pentamerids abound in the Silurian and Devonian;
Camarophoria endures to the Permian, after which the
whole sub-order is extinct.

ORDER IV. : TELOTREMATA.

Articulate Brachiopoda in which the delthyrium is re-
stricted by a pair of deltidal plates ; the brachia are
supported by calcareous crura, loops or spirals.

SUB-ORDER i. Rhynchonellacea. — Crura-bearing
Telotremata. Shell fibrous, not punctate. Deltidial
plates more or less embracing foramen, which is small,
often elliptical, and does not encroach on the umbo
(hypothyrid). Typical form short and stout (length,
breadth, and thickness not very unequal) with short
curved hinge-line, usually with a fold on the dorsal valve
and sinus on the ventral ; smooth or ornamented by fine
radial lines or strong radial ribs, sometimes by spines.
Teeth supported by dental plates. Cardinal process
usually wanting.

The rhynchonellids can generally be recognized by
their form, though Tnplesia among strophomenids and
Camarophoria among pentamerids are similar. The
presence of a sinus in the ventral valve and fold in the
dorsal, with consequent V or W shaped plication of the
valve margin, is very generally characteristic of this sub-
order ; it is one method of separating the inflowing and
outflowing water-currents.

In some cases they approach terebratulids in shape,

36 PAL/KONTOLOGY

but from them they can be distinguished by the im-
punctate shell, the surface of which, under a lens, com-
monly shows a silky fibrous appearance. At one time
all the species of the sub-order were included in the one
genus Rhynchonella, and although the Palaeozoic and
Triassic forms have been separated off into separate

FIG. 8. — PALAEOZOIC RHYNCHONELLIDS.

a, a', \Vilsonia wilsoni (J. Sowerby), xf; b, b' , Pngnax acuminatus
(Martin), x £ ; a, b, side-views ; a', b' , anterior view.

genera, the sorting of the Jurassic and Cretaceous species
is only begun.

Among Palaeozoic genera are Camarotachia (Sil.-Carb.)
with a small spondylium and septum in the ventral valve ;
Wilsonia (Fig. 8, a ; Sil.) similar, but of more globose
shape and finer ribbing ; Rhynchotreta (Sil.) with acumi-
nate beak on which the foramen encroaches, and con-
spicuous deltidial plates; Pugnax (Dev.-Carb.) with very
strong fold and sinus, and few strong ribs or even smooth
as in the familiar Carboniferous Limestone species,
P. acuminatus (Fig. 8, b).

THE BRACHIOPODA 37

Among Mesozoic genera the true Rhynchonella (R. loxia)
is rare, being only known from the Upper Jurassic of
Russia : it is not unlike Pugnax acuminatus in form. Very
similar in form, but differing in being smooth instead of
finely striate, are the Middle Lias species acuta (now
named Homceorhynchia) and the Inferior Oolite cynocephala.

Somewhat similar, but with more numerous ribs,
originating rather suddenly after a long smooth stage, is
Tetvavhynchia (R. tetraedra) of the Middle Lias.

Another very characteristic species-group, not yet
given a generic name, is that of " Rhynchonella " concinna
of the Bathonian, in which the fold and sinus are very
slight, the ribs close-set and rather fine, and the beak

b

FIG. g. — DELTIDIAL PLATES OF RHYNCHONELLIDS.

a, Rhynchonella obsoleta, Davidson (non Sowerby), x 2 ; b, R . concinna
(}. Sowerby), young, X2; d, deltidial plate. (Original.)

very acute and prominent. Young specimens of this
group are the best on which to study the deltidial
plates (Fig. 4, c), which in so many rhynchonellids
become obscured by the curvature of the beaks. Acantho-
thyris has numerous fine ribs ornamented with spines : it
is a Jurassic genus, but a similar spiny rhynchonellid
lives off the east coast of Japan.

The commonest Cretaceous species belong to the
genus Cyclothyris : they are rather broad forms, with very
fine and close-set ribs, distinct fold and sinus, and delti-
dial plates protruding mere or less as a tube around the
pedicle. Such are C. latissima of the Lower Greensand,
C. grasiana of the Upper Greensand, and C. plicatilis of
the Chalk.

38 PALEONTOLOGY

SUB-ORDER 2. Terebratulacea. — Loop-bearing Telo-
tremata, with punctate shell, foramen encroaching on
the umbo (epithyrid), and deltidial plates not bounding
the foramen laterally. Common form, ovoid and smooth ;
sometimes with a few coarse plications, consisting of a
radial elevation (fold) on one valve, and a corresponding
depression (sulcus or sinus) on the other; very rarely with
numerous ribs. Ventral umbo and foramen usually much
larger than in rhynchonellids.

The terebratulids first appear at the opening of the
Devonian period. One of the Devonian genera has so
many peculiarities that it is best described apart from
the bulk of the sub-order.

Stringocephalus buvtini of the Middle Devonian is a
large shell, nearly circular but for its straight hinge-line
and high and pointed (rostrate) ventral umbo. Unlike
other terebratulids it is hypothyrid, and the deltidial
plates arch over and meet above the foramen. The shell
does not show the typical terebratulid pattern of puncta-
tion. The loop is long and wide, parallel to the margin
of the dorsal valve ; and there is a median septum in
each valve. The cardinal process is so long that it has
to bifurcate to avoid the ventral septum.

The remainder of the. terebratulids are divided into
short-looped forms with obsolete dorsal septum and rarely
with dental plates, and long-looped forms with strong
dorsal septum, and in most cases dental plates.

The short-looped forms begin with the Devonian
Centronella, smooth, with a very simple loop, not
doubled back like that of later genera. Dielasma (Dev.-
Carb.) has a doubled-back loop, large crural lamellae
running towards one another from the top of the
descending lamellae, and strong dental plates. Similar
forms without dental plates are common in the Jurassic
and Cretaceous, and are usually placed in the genus

THE BRACHIOPODA 39

Terebratula, but that genus in the strict sense is Cainozoic
only. A few well-marked Mesozoic forms have been
given separate generic names. Thus Terebratulina has a
ring-like (annular) loop, and is finely ribbed ; Glossothyris
has a very small loop and an anterior dorsal sulcus ;
Epithyris (Fig. 5, a) has a small dorsal septum, a rather
large loop, a rather unusual beak, and more or less of a
quadriplicate margin ; Pygope and some other genera
pass through a Glossothyris -like stage, and the lateral out-
growths meet and unite again around a central perfora-
tion (these are found in the transition beds from Jurassic
to Cretaceous in the Alpine region) ; Liothyrina (Cret.-
Rec.) has a number of radial grooves in the inner surface
of the ventral valve (well seen in internal casts).

Although long-looped forms are known at least from
the Trias, it is in the Lias that the simplest form
first appears. This is the minute Zellania, in which the
loop is incomplete and the dorsal septum anterior in
position. This is followed by Megathyris, a broad form
with a loop not unlike that of Stringocephalus, but sup-
ported by two lateral septa as well as the median
septum. Several allied genera are known in the sea to-
day, survivors of this primitive group. The higher
recent genera fall into two families, one in northern, the
other in southern seas, and the highest in each case
(Dallina, northern; Magellania, southern) has a very
similar adult loop, but they attain it respectively through
two quite distinct series of metamorphoses, and the
temporary loop-forms of these genera correspond to adult
loop-forms of other genera, mostly extinct. These facts
are illustrated in Fig. 9. It will be noticed that some
of the intermediate stages are represented by recent,
not fossil, genera. This can be accounted for by the
imperfection of the geological record, for the genera
in question are small forms, not easily preserved as

PALAEONTOLOGY

MAGELLANIN/t
(Austral)

Magellania
(Cainoioic -Recent)

Terebratella
(Jurassic -Recent)

Meaerlina
(Recent)

MEGATHYRIN/C

Megath/ris [Argiope]
(Jurassic — Recent]

Cislella
(CreUceous-Recentj

DALLININ/E

(Boreal)

Dallina
(Cainozoic-Reccnt)

Trigonosemus
(Cretaceous)

Muhlfeldtia
(Jurassic-Recent)

Ismenia
(Jurassic)

Platidia.
(Recent)

Zellania
(Jurassic)

FIG. 10. — EVOLUTION OF THE LOOP IN LONG-LOOPED
TEREBRATULACEA.

Loop in outline, septum black. (After Beecher.)

THE BRACHIOPODA 41

fossils, and not easily found by collectors when they are
preserved.

In addition to the genera illustrated in Fig. 10, the fol-
lowing are noteworthy :

Ornithella (Figs. 2, 5, b) of the Jurassic, smooth, with
straight anterior margin; Aulacotkyris, Jurassic, smooth,
with ventral fold and dorsal sinus ; Lyra and Trigonosemus,
Cretaceous, plicate forms, with high cardinal areas.

SUB- ORDER 3. Spiriferacea. — Telotremata with cal-
careous spirals ; deltidial plates in some cases fused into
a single plate.

Spirifer striatus (Fig. 7, k) of the Carboniferous Lime-
stone is a very broad shell, being just about twice as broad
as long. The greatest breadth is at the hinge-line, which is
straight. Both valves are strongly convex, but the ventral
has a deep median sinus, the dorsal a corresponding fold.
The surface is ornamented with numerous radiating costae
The umbo of the ventral valve is prominent, but not
large, and below it is a large cardinal area showing
growth-striations at right-angles to the hinge-line ; it
includes a triangular delthyrium, in which a large pseudo-
deltidium arches over the pedicle-foramen. In the dorsal
valve there is a much narrower area.

The internal structure can only exceptionally be made
out. The chief feature is the pair of spirals. Starting
from the crura the ribbons first approach one another,
and then, as the coiling begins, diverge from one another,
so that the apices of the spiral cones are near the outer
ends of the hinge-line. From this it would seem that,
unlike most other brachiopods, Spirifer had a median
inhalent current and a pair of lateral exhalent currents.
The first turn of one spiral is joined to that of the other,
on the dorsal side, by a simple bar, the jugum. In the
ventral valve there are short dental plates.

Various other species of Spirifer will answer this same

42 PALEONTOLOGY

description very closely, e.g., S. bisukatus of the Carbon-
iferous Limestone, or 5. verneuili of the Devonian. The
genus in the strictest sense ranges from Devonian to
Permian, but there is. an allied form, Delthyris, small and
coarsely plicate in the Silurian, e.g., D. crispus of the
Wenlock Limestone.

Other genera closely allied to Spirifer are — Martinia
(Carb.-Perm., without dental plates and with tendency
to smoothness of surface), Syvingothyvis (Upper Dev.-
Lower Carb., with very large ventral cardinal area,
facing in a direction at right-angles to the plane
separating the valves, instead of parallel to it as usual,
and a peculiar "split-tube" or syrinx between the dental
plates), and Cyvtia (Fig. 7, /, Sil.-Dev., with high ventral
area, and narrow delthyrium with pseudodeltidium per-
forated centrally for the pedicle). Not quite so near are
Cyrtina (Sil.-Carb.) and Spiviferina (Carb.-Jur.) with
punctate shell and ventral median septum supporting
dental plates : they resemble in form the previous genera
which they respectively resemble in name. All these,
and various other genera, are usually united into one
family Spiriferida.

In the family Atvypida we find the only Ordovician
spire-bearers, such as Zygospira, in which not only are the
spirals short and simple, but their apices are directed
towards one another, so that they must have had a
median exhalent current as in normal brachiopods.
Thus in every respect they are the most primitive spire-
bearers. The later (Sil.-Dev.) and more familiar Atrypa
(Fig. 7, m) has the spiral cones parallel with dorsalward
apices.

Lastly, the family Athyrida includes the forms with
most complex brachial skeleton. The spirals point away
from one another as in Spirifer, but they start by a sharp
bend back from the crura, and the jugum is never a

THE BRACHIOPODA 43

simple bar, as in the two other families : its least complex
form is a Y, but it may be scissors-shaped or have still
more elaborate forms. Genera in this family usually
have a rounded form and short hinge-line. Meristina is a
Silurian genus, with a median dorsal septum, and paired
ventral septa (the converse of the contemporary pen-
tamerids); Athyris and Semimda are Carboniferous forms*
the latter mimicking the contemporary Dielasma, but dis-
tinguishable from it not only by its internal spirals, but
also by its non-punctate shell.

The Spiriferacea appear in the Ordovician, where only
primitive atrypids are found ; all families are represented
from the Silurian to the Carboniferous, when the atrypids
die out. The other two families survive to the earlier
part of the Jurassic period, some -species of Spiriferina
being quite common in certain zones of the Lias. The
Whitbian age (time of lower part of Upper Lias) seems
the date of the final extinction of all Palaeozoic types
of brachiopods — not only spiriferids (Spiriferina) and
athyrids (Koninckella), but also leptaenids (Cadomella)
making their last appearance. Certain minute forms in
the Inferior Oolite that have been ascribed to the
spiriferids really belong to other groups (Megathyrinae
and Rhynchonellids).

Short Bibliography.

This and succeeding short bibliographies are intended
to give the student some idea of the original works from
which he may obtain more detailed information than can
be given in any textbook, and of which he must study
those which are relevant if he is engaged in research —
even the simple research involved in naming accurately
the fossils he collects. In selecting works for these
lists, regard has been had to accessibility (to English

41 PALEONTOLOGY

students) as well as to importance. Reference should
also be made to Chapter XII. and the bibliography
there given.

BUCKMAN, S. S. — (i) Brachiopoda of Namyau beds,
Mem. Geol. Surv. India, n.s., vol. iii., mem. 2 (1917).
(2) Papers on Brachiopod homceomorphy, Quart. Journ.
Geol. Soc. London, 1906-07-08.

DAVIDSON, T. — British Fossil Brachiopoda, Palteonto-
graphical Society, 6 vols. (1851-86). The last volume
is a full bibliography of Brachiopoda up to date of
publication.

HALL, J., AND CLARKE, J. M. — (i) Palaeozoic Brachio-
poda, Palceont. of New York, vol. viii. (1892-95). A fun-
damental work on classification, summarized in (2)
Introduction to the Study of the Brachiopoda, Rep. New
York State Geologist, pts. i and 2 (1892-93).

OEHLERT, D. P., in Fischer's Manuel de Conchyliologie
(1887).

REED, F. A. C. — Brachiopoda, in Cambridge Natural
History (1895).

THOMAS, I. — (i) British Carboniferous Orthotetinae,
and (2) British Carboniferous Producti, Mem. Geol. Surv.
Great Britain : Palceont., pts. 2 (1910) and 4 (1914).

VAUGHAN, A. — (i) The Carboniferous Limestone
Series (Avonian) of the Avon Gorge, Proc. Bristol Nat.
Soc., ser. 4, vol. i. (1906) ; (2) Various papers in Quart.
Journ. Geol. Soc. (1905-15).

WAAGEN, W. — Salt Range Fossils: Brachiopoda,
Palaontologia Indica, ser. 13, vol. i., pt. 4 (1882-85).

WALCOTT, C. D. — Cambrian Brachiopoda, U.S. Geol.
Surv. Monograph, 51 (1912).

II

THE LAMELLIBRANCHIA

ASSOCIATED with the terebratuloids in the Cornbrash,
though for the most part less well preserved, are other
bivalve shells obviously different from brachiopods.
In many of them the two valves are obviously right
and left, not dorsal and ventral, since they are like
mirror - images of one another ; and though in some
the two valves are unequal, they do not show the
perfect symmetry of brachiopod-valves. There is never
a pedicle-perforation, and though there is a hinge-line
and often a cardinal area not unlike those of brachiopods,
the internal muscle-impressions are very differently
arranged, and there is never a brachial skeleton.

These bivalves belong to a distinct phylum from the
Brachiopoda — the great phylum Mollusca, of which they
constitute a class, the Lamellibranchia. The resemblances
between them and brachiopods is due to their leading a
similar life and having the same needs ; the differences
are due to the fact that they had a different ancestry, so
that a different initial structure had to be adapted to the
same needs.

As in brachiopods, each valve has an umbo and near it
the hinge-line, but the anatomy of the animal shows that

45

46 PALEONTOLOGY

these structures mark the dorsal region, not the posterior,
as in brachiopods. The region where the valves separate
most widely when the shell opens is therefore ventral
(instead of anterior). The measurement from the dorsal
to the ventral edge is the height of the shell, the measure-
ment at right angles to this (in the plane separating the
valves) is the length, being from the anterior end to the
posterior; and the measurement across both valves, at
right angles to both height and length, is the thickness.

Lamellibranchia are very rare in the oldest strata ; they
were rarer than Brachiopoda throughout the Palaeozoic
era ; as the latter diminished in numbers, they increased,
until in the Cainozoic era and at the present time they far
outnumber the Brachiopoda. They are less restricted
in their distribution, a few genera being found in fresh
waters, though the great majority are marine.

Though the mode of life is similar to that of brachio-
pods, they are more free to move about. A few fix them-
selves after the larval stage by cementation (as the
oysters) or by silky threads (byssus), as the mussels, never
by a muscular pedicle ; many burrow in sand or mud,
a few bore into harder materials, the majority move
sluggishly about the bottom, and a very few (as the
scallops) swim by a series of jumps.

In the majority of cases, the animal is bilaterally
symmetrical, and the right and left valves are as nearly
counterparts as is possible, seeing that the hinge-teeth of
the one must come opposite the sockets of the other
(equivalve shells). But in some families, especially those
with a fixed habit, the valve on which the animal lies is

THE LAMELLIBRANCHIA 47

the larger. This may be the left valve, as in the oyster,
or the right (ineqiiivalve shells).

i. Pectunculus. As a first example of a lamellibranch
we will take this genus, one very common at the present day
and in the Cainozoic era. Familiar species are P. glycimeris
of the Red Crag, and P. deletus of the Barton beds. The
shell is circular in outline and is lenticular in shape, the
valves are symmetrical to one another and very nearly
symmetrical in themselves, the distinction between
anterior and posterior ends being very slight. In fact
the only external distinction in most species is that the
umbo is very slightly nearer to the anterior end and faces
towards the posterior end, and these facts would not help
us much in deciding which end was which, for although
in most lamellibranchs the umbo is nearer the anterior
end, yet there are exceptions ; and on the other hand
the umbo usually points towards the anterior end.

The external surface is marked by radial and concentric
lines : the latter are always very slightly marked, the
former vary in different species from very faint striae to
coarse ribs.

Beneath the umbo and above the hinge-line there is
an obtusely triangular cardinal area, which in a fossil
shell is seen to be bounded above by a slightly raised
margin, and to bear a number of ridges arranged like
a set of inverted ' Vs,' one within the other. In a living
animal this area is concealed by a brown leathery mass
uniting the two valves, known as the elastic ligament.
(Occasionally this is preserved in fossils.) This elastic
ligament is something unknown in Brachiopoda, and

48

PAL.EONTO LOGY

connected with a fundamental difference in the method
of opening the shell in the two groups. The lamellibranch
has no divaricator muscles : when its adductors contract
so as to close the shell tightly, the elastic ligament is
subjected to tension ; when they relax, the elasticity of
the ligament draws the two cardinal areas towards one
another and the valves open. Throughout life, however,
the adductors are never completely relaxed, and the
ligament is always more or less stretched, so that it -is

FIG. ii.— PECTUNCULUS GLYCIMERIS (LINNE), PLIOCENE.
Exterior of right valve. (Natural size, not full grown.) (Original.)

not until after death that the ligament has full play and
the valves then yawn apart more than ever during life.

Fossil lamellibranch shells are most commonly found
in one of three conditions, according to the rapidity of
their burial after death, (i) When quickly buried (or
when the animal lived and died in a burrow), the
valves remain tightly closed, the weight of the sediment
counteracting the tendency of the ligament to open them :
in this case the interior is filled with calcite or other

THE LAMELLIBRANCHIA 49

material deposited from solution (except in those genera
Whose valve-margins do not meet for their whole length)..
(2) If the muscles decay away before burial, the valves
yawn open, and sediment drifting in prevents their
closing, but they remain united. This case is well
illustrated by the specimens of Pectunculus brevirostris that

FIG. 12. — PECTUNCULUS GLYCIMERIS (LINNE), PLIOCENE.

Interior of left valve. (Natural size.) From above down are seen in order
— umbo, cardinal area, curved row of teeth, edge of hinge-plate,
adductor impressions, pallial line, crenulate ventral margin. (Original.)

abound in the Bognor rock, a band of hard sandstone in
the London Clay of the Sussex coast. (3) If burial is
long delayed, the valves are drifted about until not only
the muscles but the ligament also is decayed ; then the
valves become completely separated. This is the case
with the abundant valves of P. glycimevis in the Red
Crag of Suffolk. Each of these three conditions may be

4

50 PALEONTOLOGY

represented by internal and external casts, the material
of the internal cast in the two latter cases being
rock-matrix, in the first case usually calcite or calcium
phosphate, etc.

On examining the interior of a valve of Pectuncnlus, we
see below the cardinal area a curved hinge-plate bearing
a row of teeth, all practically alike, usually slightly
V-shaped and about twenty or more in number. Com-
parison of young and old shells of the same species
shows that as growth proceeds new teeth are added at
each end of the row, while the cardinal area as it in-
creases in size encroaches on the central teeth until they
become obliterated.

Around the margin of the ventral half of the valve
is a series of radial grooves and ridges, by which one
valve locks into the other when tightly closed. These
look very much like a continuation of the row of hinge-
teeth, but are separated from it at either end by a short
smooth area. Margins provided with such ridges and
furrows are said to be cvemdate.

The adductor muscle-impressions are two in number,
anterior and posterior in position. They differ slightly
in shape and size, more in some species than in others.
From the inner margin a very distinct line runs up to the
umbo, marking off the path of shifting of each muscle as
the growth of the shell proceeded.

From one adductor to the other there runs, parallel to
the ventral margin, a very distinct line, the pallial line.
This takes its name from the mantle or pallium, the two
halves of which line the interior of the two valves, much

THE LAMELLIBRANCHIA 51

as in brachiopods. The edges of these mantle-folds are
thickened, both because this is the region by which new
shell is secreted, and because there is here an important
muscle (orbicular muscle) by which the lips of the mantle
can be pressed together even when the valves are not
closed. The pallial line marks the inner margin of this
thickening of the mantle. The mantle-edges separate
only at certain points — (i) antero-ventrally, where the
foot may be protruded ; (2) posteriorly, where they enclose
two openings like a figure 8 : the lower opening being
that by which water is sucked in (inhalent aperture), the
upper that by which it is expelled (exhalent). These two
apertures are in some lamellibranchs extended into long
projecting siphons, and in that case the necessary local
increase in the orbicular muscle, to form a retractor for
the siphons, causes the pallial line to be more or less
indented as it approaches the posterior adductor. Such
an indentation is termed a pallial sinus : it is not present
in Pectuncuhis.

It is necessary now to point out which of the characters
seen in the shell of Pectuncuhis are not characteristic of
all lamellibranchs, and for these technical terms must be
given.

The shell of Pectunculus, having the umbo directed
towards the hind end, is opisthogyral. The elastic ligament,
as it lies entirely dorsal to the hinge-line, is an external
ligament or tension-ligament. As it extends without inter-
ruption fore and aft of the umbo, it is amphidetic. The
same adjective may be applied to the cardinal area. The
hinge, being set with numerous undifferentiated teeth, is

52 PALEONTOLOGY

taxodont. As there are two adductor muscles, the shell is
dimyarian ; and, as these two are approximately equal in
size, it is isomyarian. As there is no pallial sinus, it is
integripalliate.

2. Nucula is a genus represented by several hundred
species ranging from Silurian to Recent, and varying
very little in that long period. The following description
applies specially to the Pliocene and Recent species,
N. nucleus, but in all essentials it will apply to any other.
(Fig. 13).

The shell is oval tending to triangular, equivalve, the
surface (as preserved in the fossil state) nearly smooth
with very delicate concentric striae. The umbones are
opisthogyral, and situated much nearer the posterior
end, so that without a knowledge of the living animal
the posterior end would be taken for anterior. There is
nothing like the large dorsal area of Ptctunculus, but the
region just behind the umbo is slightly flattened or con-
cave, and in some species forms a large depressed area
called an escutcheon. Internally, the hinge is taxodont, there
being a narrow hinge-plate bearing a long row of simple,
slightly curved teeth. Just below the umbo this row is
bent and almost interrupted, and the hinge-plate extends
downwards in a projection that bears a deep hollow
facing its fellow in the other valve. These hollows are
ligament-pits, the representative of the elastic ligament
running across the median plane from one to the other.
Such a ligament, lying below the hinge-line, is described
as an internal ligament, or better as a resilium, for when the
valves are closed it is under compression, not under

THE LAMELLIBRANCHIA

53

tension like the external ligament of Pectunculus, and it is
its resiliency (or tendency to recover its shape when the
pressure is removed) that causes the valves to open.
Nucula is isomyarian and integripalliate ; its valve-
margins are slightly crenulate. An important difference
from Pectunculus is that the inner shell-layer is nacreous
(pearly), that is to say it is composed of thin oblique
laminae of calcite bound together by organic matter

ant.adJ.

FIG. 13.— NUCULA.

Left-hand figure, internal view of right valve of N. margaritacea, Lamarch,
Eocene, (x 2. ) (After Deshayes.) Right-hand figures, dorsal view
(above) and right side view of internal cast of N, pectinata, J. Soweiby,
Albian (Gault). (Natural size.) (After Woods.)

(conchiolin), and the reflection of light from the
surfaces of such thin laminae results in interference and
consequent iridescent colours.

3. Trigonia is a genus very abundantly represented by
well-preserved specimens in the Jurassic system, to a less
extent in the Cretaceous, and very rare in later formations,
though still surviving in Australian seas. We will choose
the Upper Jurassic species Trigonia bronni (Fig. 14) for

54 PALEONTOLOGY

description, though most that is said, apart from details of
shape and ornament, will apply equally to other species.
The shell is equivalve, the shape between oblong and
triangular, the posterior end truncated, the height about
two-thirds the length ; the umbo is near the anterior end
and opisthogyral. The greater part of the surface is
ornamented with tubercles, arranged in curved rows
oblique to the margin and tending to run together
anteriorly ; fine growth-lines can be traced over the
surface of the tubercles where these are not too worn ;
but from the umbo to the truncated posterior end there
extends an area in which the ornament is different, the
concentric striae being more marked, the tubercles much
smaller and in three radial rows. As it is the edge of
the mantle which secretes the shell, this difference in
ornament depends on a difference in the secreting activity
of the siphonal part of the mantle from that of the rest of
the mantle. Enclosed between this siphonal area and
the dorsal margin is a narrow segmental escutcheon, not
extending to the posterior end, and bearing only striae
whose obliquity to the margin shows that the shape
of this area has remained the same during growth. In
its extreme anterior portion, close to the umbo, is the
short but thick external ligament. This may be preserved
in the fossil ; if not, the position it occupied can easily be
recognized, when the two valves are in position, by the
gap between their margins. Being entirely behind the
umbo, the ligament is described as opisthodetic (in contrast
to the amphidetic ligament of Pectunculus.) Although
external, it does not act quite in the same way as that of

THE LAMELLtBRANCIIIA

55

56 PALEONTOLOGY

Pectunculus. It is shaped like a cylinder slit open along
one radius, and yawning at the slit. The closing of the
valves tends to close this fissure, and the outer layers
are under tension while the centre is compressed.

The interior of the shell is nacreous ; the wear and tear
of the outer layer in the region of the umbo often leads
to the exposure of the nacreous layer on the exterior.

The hinge-teeth are quite different from those of
Pectunculus : they are very large and few in number, two
in each valve (the left posterior one bifid). Their surfaces
of contact are deeply grooved, as is also the posterior
surface of the left posterior socket (this is sometimes taken
as a third tooth, but it does not project beyond the median
plane as a tooth must do). The right anterior and the left
anterior and part of the bifid tooth are carried on a kind of
buttress (myophoric lamina) which also supports the anterior
adductor muscle. This muscle - impression is partly
crossed by a series of ridges (which probably give greater
" key " or grip to the muscle). The posterior adductor
is rather larger, and is placed nearly half-way between
the posterior end and the umbo, just behind the posterior
hinge-teeth. The pallial line has no sinus. The margins
are not crenulate.

Thus Trigonia, Pectunculus, and Nucula are all three
equivalve, opisthogyral, and -integripalliate, and Trigonia
and Nucula agree further in having a nacreous interior,
and in being very inequilateral (though in Trigonia it is
the anterior region, in Nucula the posterior, which is the
shorter). The opisthodetic ligament and hinge-teeth are
features in which Trigonia differs strikingly from both the

THE LAMELLIBRANCHIA 57

other genera. This type of hinge is described as schizodont.
According to the method of notation to be explained

immediately, its formula is — — •

4. Cyrena semistriata is an abundant fossil in the
Oligocene beds of the Isle of Wight. The shape is sub-
trigonal, the umbo being about one-third the total length
back from the anterior end ; the anterior border is rounded,
the posterior part of the dorsal border sloping obliquely
back to a narrow rounded posterior end. (Apart from

FIG. 15. — CYRENA SP. RECENT.

Interior of left valve. (Natural size.) Teeth white, sockets black.
p.s., Pallial sinus ; tig., ligament. (Original.)

this shape, the remainder of the description will apply
in general to any other species of Cyrena, such as that
shown in Fig. 15.) The shell is equivalve. The outer
surface is smooth, with concentric growth-striations.
The ligament is opisthodetic and short. Internally, the
shell is isomyarian, and there is a very slight pallial sinus,
so that it may be described as feebly sinu-palUate. The
hinge-teeth are carried on a well-defined hinge-plate
and fall into three distinct sets (heterodont) : just under
the umbo are a series of short, more or less vertical

58 PALEONTOLOGY

teeth, radiating from the umbo — the cardinal teeth ; in
the anterior and posterior regions of the hinge-plates there
are long teeth, more or less horizontal, parallel to the
shell-margin — these are called (inaccurately) the lateral
teeth. The anterior-lateral teeth come close up to the
cardinals, but the posterior-laterals are separated from
them by a space. The posterior-laterals are always
posterior to the ligament— this fact is the essential dis-
tinction between them and the cardinals, the posterior of
which may in many heterodont shells be long and nearly
horizontal, so that it might be mistaken for a posterior-
lateral if its position relative to the ligament is not noticed.
To come to details — in the right valve there are two
laterals, the inner of which is on the inner margin of the
hinge-plate, while the outer one is not quite on its outer
margin ; there are three cardinals, of which the middle
one is bifid ; and two posterior-laterals, occupying similar
positions on the hinge-plate to the two anterior-laterals.
In the left valve, there is a well-marked anterior-lateral
tooth, not quite on the inner margin of the hinge-plate
(since it has to interlock between the two anterior-laterals
of the right valve) ; the outer margin of the hinge-plate
is very slightly raised into a vestigial tooth. There are
three cardinal teeth, each fitting in behind the correspond-
ing tooth of the right valve. The posterior-laterals are
similar to the anterior-laterals.

The general arrangement of the teeth is always the
same in all Heterodonts : the lateral teeth of the left
valve are always on the outer side of (i.e., above)
the corresponding teeth of the right valve. This

THE LAMELLIBRAXCHIA

59

makes possible a very convenient system of symbols,
for which we are indebted to the French palaeontologists
Munier-Chalmas and Bernard. The lateral teeth are
numbered from within outwards, so that those of the
right valves have odd numbers, those of the left even :
thus the right anterior cardinals are Ai. and Am., the

FIG. 16. — HETERODONT HINGES.

Diagrammatic representation of Cyrenoid type (above) and Lucinoid type
(below). Interior views of right valves, all possible teeth represented,
and those originating from the same lamina shown as connected.
Teeth of right valve, black ; sockets corresponding to teeth of left
valve, dotted.

left An. and Aiv. ; the posterior-laterals Pi. and Pin.,
PH., and Piv. The symbols for the cardinals are
based upon the facts of early development, which show
them to be the detached hook-like ends of laminae of which
the main portion forms the anterior-laterals : thus the
right middle cardinal is derived from lamina i., and so is

6o PALEONTOLOGY

symbolized by the arable numeral i ; the other two are
derived from lamina in., and are called 30 (the anterior)
and 36 (the posterior). Similarly in the left valve
the anterior and middle cardinals are both derived from
lamina n.,so they are called 20 (anterior) and 2b (middle),
while the posterior comes from lamina iv., and so is
called 46. These facts are expressed diagrammatically
in Fig. 1 6. It will be noticed that the alternation of
the cardinal teeth follows that of the anterior-laterals
from which they are derived — those numbered 2 fit in
between those numbered i and 3 ; those numbered 3
between those numbered 2 and 4.

Heterodont hinges like the one described are known
as the Cyrenoid type ; there is another type with
fewer teeth, the Lucinoid type, shown diagrammatically
in Fig. 1 6. In this, the central tooth of the hinge is in
the left valve — it is 2, instead of i. The complete dental
formula for each may be written thus :

Lucinoid type ...» ,-

Cyrenoid type .

A shorter formula, omitting the laterals, gives -W

2, 40

for the Lucinoid and ' ^a' , for the Cyrenoid type.
20, b, 46

A dentition rarely contains the full set of teeth, but it
can still be classed as Lucinoid or Cyrenoid ; the laterals
in particular are often reduced in numbers, and when
the umbo is far forward the anterior-laterals may be

THE LAMELLIBRANCHIA 61

crowded out altogether. The posterior-laterals, again,
are a late development, resulting from the shortening-up
of the opisthodetic ligament : in early forms with a long
ligament they are wanting.

Cyrena is a freshwater lamellibranch, ranging from
Jurassic to Recent. Modern species live only in sub-
tropical and tropical streams and mangrove-swamps, but
the genus (including the sub-genus Corbicnla, in which the
lateral teeth are cross- striated) survived in Britain until
well on in Pleistocene time.

5. Crassatella sulcata is a common fossil in the
Barton Clay (Upper Eocene) of the Hampshire Basin.
In shape it is much like Trigonia, but the umbones
are forwardly directed (prosogyval). The surface is
ornamented with strong concentric ridges, but many
other species of the genus are nearly smooth (e.g., the
large C. plumbea of the Paris Basin, Fig. 17). There
is an escutcheon, and in front of the umbo there
is a very similar, but smaller, depressed semicircular
area, the lunule. Internally the shell is not nacreous ; it
is isomyarian and integripalliate ; the valve-margins are
crenulate (though not in all species of the genus). The
hinge-plate is very narrow in front and behind, but very
deep in the umbonal region, where it bears a ligament-
pit, much larger in proportion than that of Nucula. In
front of this are two nearly-vertical cardinal teeth :
those of the left valve (20, zb) are both large, those of the;
right valve fit in front of those of the left, and the anterior
tooth (3#) is very small. The articulating surfaces of
the teeth are slightly cross-grooved like those of Trigonia^

62

PALEONTOLOGY

Both anterior and posterior laterals are present, one of
each in each valve. (In C. plumbea both 30 and the
laterals are obsolete — i.e., so reduced as to be scarcely
recognizable.) As the right anterior-lateral fits under the

re s i lvu/m>

FIG. 17. — CRASSATELLA PONDEROSA (LINNE). (x£.)

Right valve above, left valve below. Teeth white, sockets black.
(After Deshayes.)

left, while the right posterior fits over the left, the teeth
are probably Ai. and An., Pin. and Pn.

Crassatella, at the present day, is confined to the
Indo-Pacific province and the tropical parts of the

THE LAMELLIBRANCHIA 63

Atlantic. It. is known fossil from the Lower Cretaceous
onwards, and was less restricted in its geographical range
until after the Eocene, as is shown by its presence in the
Barton Clay and (much more abundantly) in the Eocene
of the Paris Basin.

6. Meretrix (Sinodia) incrassata (Fig. 18) is an
abundant fossil in the " Venus-bed " of the Lower
Oligocene of the Isle of Wight and Hampshire — that
bed taking its name from the genus in which this species
was at one time included.

FIG. 18. — MERETRIX (SINODIA) INCRASSATA (J. SOWERBY),
OLIGOCENE.

Interior of right valve. (Natural size.) (Original.)

The shell is rounded and sub-triangular, prosogyral,
with fairly well marked lunule, without escutcheon, but
with a fairly long opisthodetic ligament. The surface
is ornamented with concentric striae. The hinge-plate
is large and thick, and bears three cardinal teeth
in each valve, with two unequal anterior-lateral teeth
(Ai., Am.) in the right valve, and one (An.) in the left,
the hinge being of Cyrenoid type. The absence of
posterior-laterals is correlated with the length of the

64 PALEONTOLOGY

opisthodetic ligament : only in genera in which the
ment becomes shorter do posterior-laterals appear, to
strengthen the hinder region, which no longer has a
ligamentary connexion. The shell is isomyarian and
sinu-palliate t there being a well-marked pallial sinus
caused by the large size of the retractor muscles of the
long siphons. The valve-margins are sharp and not
crenulate.

Thus Meretrix is heterodont and sinu-palliate, and there-
fore representative of what is regarded as the highest
type of lamellibranch — certainly one of the most modern
types. The genus has a world-wide distribution to-day,
and species attributed to it are known as far back as the
Jurassic.

7. Pseudomonotis echinata(Fig. 19) is the best pre-
served lamellibranch found in the Cornbrash. It is small
(length about 14 mm., height 16 mm., thickness 7 mm.),
rounded, and strongly inequivalve : the left valve being
convex, with prominent umbo, and ornamented with
spiny radial ribs ; the right valve being nearly flat, with
inconspicuous umbo, and nearly smooth, with only slight
radial striae. The outline of each valve is somewhat
gibbous — oval with the axis curved in a posterior direc-
tion ventrally ; but as the outline is traced round to the
umbo posteriorly it is seen to be deflected out, giving
rise to a flat triangular projection which has the effect
of lengthening the hinge-line. Such projections are
termed ears (or wings), and a shell possessing them is said
to be auriculate (of alate). The posterior ears of Pseudo-
monotis are fairly large, the left anterior ear is barely

THE LAMELLIBRANCHIA

65

recognizable, but the right anterior ear, though small, is
quite distinct and has a well-marked notch below it.
Such a notch, as we know from recent forms, served for
the passage from the interior of the shell of the byssus (a
bunch of silky threads of conchiolin, secreted by the foot)
by which the mollusc could attach itself, temporarily or
permanently. The inequivalve character is here, as in
other lamellibranchs, associated with a fixed habit
of life.

Between the umbo and the long hinge-line there is in

FlG. 19. — PSEUDOMONOTIS ECHINATA (J. DE C. SOWERBY),

BATHONIAN. ( x 2.)

a, Posterior view ; b, right side view: the black portion is the interior of
the left valve where it overlaps the right. At the upper end of this is
seen the very small right anterior ear, with the byssal notch below it ;
between it and the elevated umbo of the left valve is the amphidetic
area of the left valve. (Original.)

each valve a long and low cardinal area, not separated into
lunule and escutcheon. As in Pectunculus, this is occupied
by an amphidetic ligament.

It is not easy to clear the interior of specimens of
P. echinata, but from allied forms we may infer that hinge-
teeth are feebly developed or obsolete, and that the shell
is aniso- or heUromyarian, there being a large posterior

5

65 PALAEONTOLOGY

adductor from which a pallial line extends indistinctly
forward and ends in a very small anterior adductor
impression. The edge of the left valve overlaps that of
-the right, and^he small portion of the interior thus exposed
iis seen to be nacreous.

The genus Pseudomonotis is found principally in the
Triassic and Jurassic rocks, but allied genera range
from the Ordovician to the Recent period. Among its
Recent allies is the pearl " oyster " of the Pacific Ocean,
Meleagrina margaritifera, pearls being pathological secre-
tions of nacre around parasites or other irritating bodies,
and mother-of-pearl being the normal nacreous layer of
the shell.

8. The genus Ostrea, the true oyster, ranges from
perhaps the Carboniferous period to the Recent. O.
ventilabvtim of the Oligocene (Fig. 20) is a convenient

species to describe, but (except for shape and ornament)

,'

many other species will answer to the description. The
oysters are fixed (sessile) forms, and consequently very
inequivalve. The left valve is fixed by cementation to some
solid object — often some other shell — and becomes thick
and convex ; the right valve is nearly flat and forms a lid
to the left. The outline (of O. ventilabmm) is nearly semi-
circular, the posterior margin from the umbo back being
nearly straight. The height (in the morphological sense)
is greater than the length (length 6 cm., height 7 cm.,
thickness 3 cm.) The shell is opisthogyral. In the left
valve, a variable area of the umbonal region is adherent
to some foreign body and shows only the impress of that
instead of its own ornamentation. Beyond this area the

THE LAMELLIBRANCHIA

67

ornament consists of coarse diverging ribs, sub-angular
to rounded in section, increasing in number by bifurcation
or by intercalation of new ribs. The whole surface has
a scaly look, due to the highly laminated structure of the
shell. The right valve is nearly smooth, except for con-
centric corrugations due to growth irregularities, but also
shows this laminated texture. In the umbonal region
there is an area of more irregular character corresponding
exactly to the area of attachment of the left valve. That

add

FlG. 20. — OSTREA VENTILABRUM, GOLDFUSS, OLIGOCENE. (x^.)

Right-hand figure shows both valves seen from the right side ; in the left-
hand figure the right valve is removed, add. , Adductor impression;
tig., ligament. • (Original.)

the surface of the foreign body should thus " show
through " a thick pair of valves is at first sight puzzling,
especially as no trace of it is seen on the interior of either
valve. The explanation is that the oyster fixed itself
when very young, both valves of the very thin shell, as
well as the animal's body between, being moulded up and
down over the irregularities of the surface to which it
was attached. As growth proceeded the shell not only
extended beyond the area of attachment, but also increased

68 PALEONTOLOGY

in thickness by the addition of internal layers, which
gradually smoothed over the irregularities of the inner
surfaces but left those of the outer surfaces unchanged.

Internally, there are no hinge-teeth ; there is a large
triangular elastic ligament, extending from the umbo to
the hinge-line. When the valves are closed, the outer
part of the ligament is probably under tension and the
inner part under compression. The central part is
thicker than the anterior and posterior, so that in the left
valve the ligament-area appears divided into three parts,
the central concave, the others flat : the whole ligament
area is marked by fine horizontal striations.

There is only one adductor muscle-impression, posterior
in position, the anterior adductor being entirely aborted
in the adult : the shell is therefore mono-myarian. The
interior is sub-nacreous, the iridescence being faint.

The external ribs give the valve-margin of the left
valve, which projects beyond that of the right, a coarse
crenulation. As this is traced towards the hinge it
becomes supplanted by a slightly deeper-set series of
crenulations, counterparts to which appear in the right
valve, together forming an approach to a taxodont denti-
tion. These " teeth " do not interlock tightly, however ;
they are only found in a few species of Ostrea ; they are
certainly not the relics of the hinge-teeth of the ancestors
of the Ostreida.

The oysters (including allied genera as well as Ostrea
itself) have the strongest shells of any lamellibranchs, in
the sense of standing wear and tear bast : they may be
washed out of their original deposit, rolled about by

THE LAMELLIBRANCHIA 69

waves or streams, and retain a recognizable form while
most of their associated fossils are pulverized. Hence
they are often found as derived fossils in strata much
younger than those to which they belong ; Jurassic oysters,
for instance, are very commonly found in the glacial
gravels of the Pleistocene period,

9. Corbula is a genus represented by various small
species in the Jurassic, Eocene and Oligocene strata of
Britain, such as C. pisum and C. revoluta of the Barton
Clay. The shell is inequivaive, the right valve being in
every way larger than the left though the two scarcely
differ in form ; both being oval and rostrate posteriorly,
with no ornament but lines of growth. The umbones are
nearly central in position and opisthogyral. The hinge-
structure is best made out on large species such as Corbula
gallica of the Eocene of the Paris Basin (Fig. 21). In the
right valve there is a sharp anterior-cardinal tooth pro-
jecting to an unusual extent towards the left valve and
slightly upwards. Behind this is a space with a resilium-
pit which instead of being vertical is horizontal, facing
downwards, and placed within the umbo; behind this
there is a long vestigial posterior-lateral tooth. In the
left valve, a deep conical socket corresponds to the right
cardinal tooth ; behind this is a conspicuously-projecting
horizontal plate (resiliophore), which carried the resilium
and underlies the resilium-pit of the right valve. The
interiors are not nacreous; the shell is isomyarian and
has a short pallial sinus with vertical front edge.

Modern species of Corbula are marine, living in
moderately shallow water, but an allied genus Erodona

PALEONTOLOGY

lives in the great rivers of South America beyond the tidal
limit, and is found in British freshwater Oligocene strata.

FlG 21.— CORBULA GALLICA, LAMARCK, EOCENE, PARIS BASIN.

(Natural Size.)

Upper figure, complete shell from left side ; middle figure, interior of
left valve; lower figure, interior of right valve. p.s.t Pallial sinus;
x, resiliophore. (After Deshayes. )

10. Pholadomya is a genus which made its first ap-
pearance in the Lower Lias, was abundantly represented

THE LAMELLIBRANCFIIA 71

by species throughout the Jurassic, became less common in
the Cretaceous, is still represented by several species in
the British Eocene beds and one in the Pliocene, but only
survives to-day in deep waters in the Atlantic Ocean and
West Indies, although many of the fossil species (even up
to the Pliocene) must have lived in shallow waters. We
may take P. fidicula of the Inferior Oolite as an example
(Fig. 22).

The shell is extremely thin, so much so that when
found as an internal cast, it requires careful examination
to determine whether any of the shell is still adherent or

FIG. 22. — PHOLADOMYA SP. RESEMBLING BUT SHORTER THAN

P. FIDICULA (J. DE C. SoWERBY), AALENIAN. ( X £.)

Left side view. (After Goldfuss.)

not. Internally it is nacreous. The shell is equivalve,
and strongly inequilateral, the umbones being far forward
and prosogyral. The valves gape at the posterior end,
the long siphons not being fully retractile ; consequently
the interior is always filled with rock-matrix and the
posterior end has a rough, broken appearance where the
matrix within joined that without. The surface is
ornamented with numerous oblique, slightly curved,
radiating ribs, dying away in the postero-dorsal area.
(In some species the ornament is more elaborate, with
fewer an.d coarser ribs and tubercles, and differing in

72 PAL/EONTOLOGY

different areas.) The ornament is shown by the internal
cast as well as by the outer shell. There is a short,
opisthodetic external ligament. In the fossil it is difficult
to determine whether teeth are present or not ; but in the
recent species there is only a vestigial tooth. The shell
is isomyarian and sinupalliate, though the impressions
are too feeble for this to be determined in many cases.

The ten genera of Lamellibranchia here described
. form a very inadequate sample of a very extensive class,
for which various methods of classification have been
from time to time proposed. Classifications are of two
kinds — morphological and phyletic. All early classifica-
tions necessarily belong to the former type, in which one
proceeds to sort out the members of a class first into
larger groups defined as precisely as possible by certain
differences of structure, then into smaller groups defined
by other differences, and so on. The important question
to be settled in framing a morphological classification is
—What characters shall we choose to define our larger
groups, and what to define each successive grade in our
subdivision ? According as the characters are well or
ill-chosen, the classification comes to be recognized as
more or less " natural " or " artificial." The characters
to be considered fall into several categories :

(i) The most external or superficial characters, such
as the shape and ornament of a shell. These are found
pften to differ greatly in a number of species which have
most other characters alike, hence they are recognized as
principally of value as specific distinctions; and only
when they are constantly associated with other characters.

THE LAMELLIBRANCHIA 73

can they be taken as part of the definition of larger
groups.

(2) Characters which are evidently directly connected
with the animal's particular mode of life — adaptative
characters. These have to be taken very cautiously, since
sometimes two animals from widely different groups may
come to resemble one another very closely (especially
when we have only the hard parts to deal with) on account
of their adaptation to similar conditions (convergence}. We
have already seen how many features lamellibranchs
and brachiopods have in common owing to general simi-
larities of habit, and cases of homceomorphy among
brachiopods have been mentioned. Among lamelli-
branchs themselves there are cases of similarity of
external form (e.g. Trigonia and Crassatella), accom-
panied by great differences in internal structure. Never-
theless in many cases adaptation to different lives does
form the basis of such fundamental divergence as to
determine essential boundaries in classification, as for in-
stance between the fishes and air-breathing Vertebrates.

(3) There are certain characters which are found to
run, with but slight modifications, through forms of
varied life-modes and varied external appearance. Such,
for instance, are the various types of hinge-teeth in
lamellibranchs. Such characters are found to provide
the safest bases for classification.

The history of taxonomy (i.e., the laws of classification)
shows that morphological classification tends more and
more away from what is regarded as " artificial "
towards what is " natural," and this tendency has meant,

74 PALEONTOLOGY

at first unconsciously, and later consciously, a tendency
towards phyletic classification — that is to say, a classifica-
tion expressing the phytogeny or evolutional genealogical
tree of the group to be classified. All morphological
classifications must, in fact, be more or less artificial,
since they proceed by definitions of structure, not of
derivation : only by becoming phyletic do they become
completely " natural." Conversely, the more natural
a classification, the more difficult is it to express it by
morphological definitions, since the most important dis-
tinctions between two diverging lineages are not fully
developed in the earlier members of the lineages, and
also they may be lost in aberrant members whose con-
nexion with a lineage is clearly shown by secondary
features not characteristic of the whole of the lineage.
Hence modern taxonomic definitions are full of such
qualifications as "usually," " in typical genera," " in all
but the earliest members," etc.

An early classification of Lamellibranchia was into
Monomyaria and Dimyaria, the latter being divided into
Heteromyaria and Isomyaria, and these last into Inte-
gripalliata and Sinupalliata. If the definitions are not
applied too rigidly, this forms a fairly natural classifica-
tion, and has the advantage to palaeontologists of being
based on shell-characters. But Heteromyaria are more
nearly allied to Monomyaria than to Isomyaria, and the
sinupalliate condition has been arrived at independently
along so many different (at least five) lines of descent, that
it cannot be used to define a single group. Another classi-
fication by shell-characters was that of Neumayr, based

THE LAMELLIBRANCIIIA 75

upon the hinge-teeth : the class was divided into orders
Taxodonta, Heterodonta, etc. This is in large measure
a natural scheme, but not entirely, as it brings together
Nucnla and Pectimculus as taxodonts, although they differ
very greatly in other respects, and associates Trigonia and
Unio as schizodonts more closely than seems really
justified.

Zoologists, however, dealing with recent lamelli-
branchs, and not having only the shell to examine, have
proposed as the soundest basis of classification the
structure of the gills. Thus a few genera such as Nucnla
have plume-like gills like those of the other classes of
Mollusca : these evidently form a primitive order (Proto-
branchia). Others, as Pectunculns, Trigonia, Pteria (Fili-
branchia), have gills transitional in character between the
last and the typical sack-like gills which give the name
to the whole class. Many others, including the re-
mainder of those so far described, have the typical
lamellate gill (Eulamellibranckia). Lastly, a few genera
have the gills modified into a simple septum across the
mantle chamber (Septibranchia).

This plan is unsatisfactory to palaeontologists. Not
only is it based on morphological characters that cannot
be discovered in an extinct genus, but it separates genera
(like Pecten and Lima) which shell-characters show to be
nearly related, and unites them with others with which
they have no near relation (Pecten with Trigonia, Lima
with the Heterodonts and Desmodonts). The develop-
ment of the typical lamellibranch gill is a process which
must have gone on independently in different lineages,

76 PALAEONTOLOGY

and the different types of gill give us grades of development
only, not lineages.

The latest proposal for classification (here adopted) is
that of Professor Henri Douville, of Paris, according to
whom the main lines of divergence of the lamellibranchs
were determined by original differences in habits of life.
There are three branches — active, fixed, and burrowing
— so that we have here what has been termed by Verte-
brate palaeontologists an adapt ative radiation. But the
members of any lineage are not rigidly tied down to one
groove ; in each lineage there are cases of re-adaptation
to one of the other modes of life, re-adaptations which
lead sometimes to a close imitation of one type by
another (convergence), but which can never obliterate
or reverse the effects of ancestral history.

I. Active Branch.

I. NUCULACEA.

Equivalye; isomyarian; internally nacreous; hinge-line
primitively taxodont ; without distinct cardinal area ;
surface rarely ornamented otherwise than by concentric
striae. Ligament external (Ctenodonta) or internal (Nucula,
Leda). Integripalliate (except Leda and Yoldia). Ordo-
vician to Recent.

Leda (Ord.-Rec., Fig. 23, a) and Yoldia differ from Nucula
in having the mantle drawn out into distinct though small
siphons. This greater development of the posterior region
results in a more central umbo, while there is a slight
pallial sinus, and in Leda a tendency to more or less
rostration (drawing out of the posterior end) : species
which show this in an extreme degree are sometimes
separated under the name Dacryomya. Nucula and Leda

THE LAMELLIBRANCHIA 77

share with the brachiopod Lingula the honour of being
among the longest-lived genera of animals.

Ctenodonta has some resemblance in form to Pectunculus,
but is far earlier in time (Ord.-Sil.).

Probably allied to the Nuculacea are two Silurian-
Devonian groups — (i) Actinodonta (Fig. 23, b), Lyrodesma,
etc., which show a change of the hinge-teeth from the
parallel series of Nucula to a radiating series; and (2)
Cardiola and Buchiola, with strong radial ornament and
apparently toothless hinge.

II. NAIADACEA.

Equi valve, strongly inequilateral ; isomyarian ; inte-
gripalliate; internally nacreous. Ligament external,
opisthodetic. Hinge-teeth lamellar. Rarely ornamented
otherwise than by concentric striae. Freshwater habitat.
Devonian- Recent.

The "freshwater mussels "are probably derived from
forms like Actinodonta by the restriction of the ligament,
various forms from the Permian of Russia showing
hinge-teeth of taxodont to radial type. On the other
hand, as early as the Devonian, there is found an appar-
ently toothless form, Archanodon ; and the hinge-structure
of the common Carbonicola of the Coal Measures (Fig. 23, c)
is imperfectly known. From Mesozoic time down to the
present day the genus Unio shows a hinge which is
transitional between the actinodont and heterodont types :

its formula may be given as — --— — — -^ (3a, 4a, and 5

20, o, 40, 0 *

are usually described as "anterior-laterals," 20, as a
" cardinal,"and the other three as " posterior-laterals " —
an incorrect term, since they extend below the ligament).

III. PR^-HETERODONTA.

Equivalve, inequilateral, isomyarian, integripalliate ;
internally nacreous or porcellanous. Hinge-teeth con-

PALEONTOLOGY

FIG. 23. — LAMELLIBRANCHS.

, Leda (Dacryomya] lacryma (I. de C. Sowerby), Inferior Oolite. Right side.
(X3-) (After Goldfuss.) b, Actinodonta cuneata Phillips, Ordovician.'
Interior of left valve. Teeth white, sockets black, (x^.) (After
Phillips.) c, Carbonicola robusta (J. de C. Sowerby). Coal Measures
Upper Carboniferous. Right side. (x£.) (After Hind.) d, Cardinia
hsteri (J. Sowerby). Right side. (x£.) d', C. ovalis (Stutchbury)
internal cast, (xj.) (After Quenstedt. ) et Astarte bore Us (Chemnitz),

tHE LAMELLIBRANCHIA 79

sisting of cardinals without anterior or posterior laterals .
Ligament opisthodetic. Ornament variable. Devonian-
Recent.

Under this name, Prof. Douville includes two groups
— (a) Trigoniacea, with nacreous interior and dental

formula 3^ (Carb.-Recent) ; and (b) Prae astartacea,

2> 4 b

with porcellanous interior and dental formula ^'— r

2, 46

(Dev.-Jur.)

The Trigoniacea are a small but important group-
Schizodus (Carb.-Perm.) and Myophoria (Trias) show a
gradual change from smooth and prosogyrous, with tooth
2 simple, to the highly-ornamented and opisthogyral
Trigonia with tooth 2 bifid. The geological and geo-
graphical distribution of this genus have already been
mentioned.

The Pva-astavtacea form a less compact and less familiar
group. It includes Cardinia (Fig. 23, d, d') of the Lias
(frequently placed in the Naiadacea from its external
resemblance to Unio, but non-nacreous and marine), with
its old-age (senile, or geronticj modification Hippopodium in
which the growth in the direction of thickness leads to a
most extraordinary shape. Other genera are Megalodon
(Devonian and Rhaetic) and Pachyrisma (Jurassic), large
forms in which the hinge is short and high, with big
coarse teeth ; and the little Tancvedia of the Jurassic.

FIG. 23, — LAMELLIBRANCHS (continued).

Pliocene. Interior of left valve. Teeth white, sockets black. (x£.)
(After S. V. Wood.) f, Diceras minor Deshayes, Upper Jurassic.
Anterior view of internal cast. (Natural size.) a, Attachment ;
b, c, grooves on cast corresponding to myophoric laminae ; rf, casts of
tooth-sockets ; L. K, left, and R.V.} right valve. (After S. P. Wood-
ward.) if, Cardiuin parkinsoni J. Sowerby, Pliocene. Interior of valve,
(xj.) (After S. V. Wood.)

80 PALAEONTOLOGY

IV. HETERODONTA.

Equi- or inequi-valve ; isomyarian ; integri- or sinu-
palliate ; internally porcellanous. Hinge-lamellae differ-
entiated into anterior-lateral and cardinal teeth, with
frequent development of posterior-laterals behind the
shortened ligament. Ligament opisthodetic, usually
external, sometimes internal (resilium). Ornamentation
varied. Jurassic to Recent.

This very large and important order falls into two
series, in each of which a number of sub-orders may be
recognized.

A. WITH TEETH OF LUCINOID TYPE : 3^_f,

2, 46

1. Cardiacea. — With radial ornament, at first (Proto-
cavdia) in siphonal region only, later on whole surface.
Umbones fairly central. Anterior and posterior laterals
well developed.

The common cockle, Cardium (Fig. 23, g), has a rounded
outline, a central umbo, with well-defined and separated
anterior-lateral, cardinal and posterior-lateral teeth ; the
whole surface is ornamented by strong radial costae. In
the Mesozoic Protocavdia radial ornament is confined to the
posterior end (i.e., to the part secreted by the siphonal
part of the mantle) the rest being concentrically marked.
(Compare the case of Trigonia.) In the Aralo-Caspian
region, from Miocene to Recent, are found Adacna and
Limnocardium, cockles adapted to a freshwater or brackish
habitat, with long siphons and a pallial sinus. Doubtfully
referred to this sub-order is Thetironia, common in the
Lower Greensand of the Isle of Wight, with a myophoric
ridge that has been mistaken for a peculiar pallial sinus.

2. Rudistes. — Highly aberrant fixed forms, probably
derived from Cardiacea. Inequivalve. Fixed either by
left valve (normal) or right (inverse) ; both valves spiral

THE LAMELLIBRANCHIA 81

in early forms, the fixed valve later becoming conical or
cylindrical, the free valve flattened and lid-like. Teeth

of normal forms . -.'-' —3- • of inverse forms . ^ ,

AII.-2 ' AII.-2, PlI.

according to Prof. Douville, An. and 2 being united
into one tooth in each case. The ligament becomes
deeply internal and finally disappears, the valves being
no longer hinged but the free valve sliding up and down.
The whole shell gradually loses all resemblance to ordinary
lamellibranchs. Upper Jurassic and Cretaceous.

In the Upper Jurassic of North-East France is found
Diceras (Fig. 23,7), having both valves twisted into a loose
spiral, and fixing itself indifferently by either valve. This
is the forerunner of a great series of the most remarkably
modified of all lamellibranchs. In the Lower Cretaceous
there are genera fixed some by the left valve, the majority
by the right: in either case the fixed valve tends to become
conical or cylindrical in shape, while the free valve takes
on the character of a lid or operculum. The genus
Requienia still has both valves spirally coiled, but the right
valve is quite flat, so that the whole shell has a strange
resemblance to a gastropod with a spiral operculum. In
the Upper Cretaceous the forms fixed by the right valve
develop into still more extraordinary forms, mimicking the
rugose corals of the Palaeozoic. Such is Hippurites, the
right valve of which forms a cylinder a foot or more
in height, of which the animal only occupies the upper-
most portion, the lower part of the cavity having been
cut off, as growth proceeded, by a series of calcareous
partitions (like the tabulae of a coral or the septa of
a cephalopod). The left valve fits on as a lid, and
the hinge-teeth and adductor muscles are strangely
modified. These hippurites grew in reefs, like corals,
and they (as well as their Lower Cretaceous prede-
cessors) are almost restricted to tropical and sub-tropical

6

82 PAL/EONTOLOGY

latitudes, only a few stragglers being found in British
strata (Toncasia lonsdalei in the Lower Greensand of
Wiltshire only ; Durama mortoni and a few others in the
Chalk). In the Mediterranean region they serve as zonal
indices, particular species having a narrow range in time,
contrary to the general rule among lamellibranchs. At
the end of the Cretaceous period the Rudistes died out.

3. Lucinacea. — Ornament concentric, sometimes
combined with radial. Trias, to Recent. Forms closely
resembling Lucina (but hinge unknown) occur in Silurian
and Devonian. The chief genera are the nearly circular,
usually thin Lucina (Fig. 24, a), with a long and narrow
anterior adductor (numerous sub-genera, Trias.-Rec.), the
stout Unicardium(Tria.s.-Cret), Corbis (Jur.-Rec.) in which
a fine radial ornament is seen in the depressions between
the strong concentric ridges, and Diplodonta (Eoc.-Rec.)
in which the teeth 2 and 3^ are bifid.

4. Chamacea. — Fixed forms, inequivalve, closely
analogous to the simpler Rudistes, but derived at a later
period probably from Lucinacea.

The only genus is Chama (Fig. 24, d, Cret.-Rec.), which is
fixed usually by the left valve (some species by the right).
The umbones are strongly prosogyral, the fixed valve
larger and the free valve lid-like. The exterior has a
scaly appearance.

5. Mactracea.— Derived from Diplodonta (Lucinacea),
with which they agree in having tooth 2 bifid, and differ
in being sinu-palliate and having the ligament internal
(resilium). Ornament concentric, striate.

Mactra (Fig. 24, c, Eoc.-Rec.) is oval in form. Ensis (the
razor-shell, Rec.) has taken to burrowing, and so has
acquired the oblong shape common among Desmodonts
(see later). It is, in fact, a homoeomorph of Solen, from
which -it is distinguished by its heterodont teeth.

THE LAMELLTBRANCHIA

84 PALEONTOLOGY

6. Tellinacea. — Sinu-palliate. "Ornament concentric
striate.

Tellina (Fig. 24, b), one of the commonest of seaside
shells, dates from the Jurassic. Donax (Eoc.-Rec.) is
opisthogyral, and has a short posterior side (like Nucnla).

B. WITH TEETH OF CYRENOID TYPE : * ' ^ ' 7 •

2a, b, 40

1. Cyprinacea. — Concentric ornament, rarely more
than striate; valve-margins smooth. Umbones anterior.

Various early Jurassic forms, too rarely found with
hinge decipherable, have been shown by Prof. Douville
to be the originators of this type of hinge. Better known
are Cyprina (Fig. 24, e), a sub-circular shell with beaks
curved forward, crowding out the anterior cardinal tooth,
abundant in the colder seas (Jur.-Rec.) ; Trapezium (Cypri*
cardia), a somewhat four-sided form (Jur.-Rec.); and
Isocavdia (Mio.-Rec.), with highly spiral umbones.

2. Astartacea. — Concentric (striate or costate) orna-
ment, never radial ; valve-margins usually crenulate.
Ligament external (except in Crassatella.) Integri-
palliate. Trias. -Recent.

The chief genera are the nearly circular Astarte
(Fig. 23, e) with numerous sub-genera, the trapezoidal
Opts (Mesozoic) and Crassatella (Cret.-Rec.)

3. Cyrenacea. — Fresh-water habitat. Concentric
striate ornament ; valve-margins smooth ; integri- or
sinu-palliate. Umbones fairly central.

Anterior and posterior lateral teeth well developed.
The chief genera are Cyvena (Jur.-Rec.) and Covbicula
(Eoc.-Rec.), already described.

4. Carditacea. — Radial ornament; valve -margins
crenulate ; umbones anterior, leading to a* lengthening
of the teeth 2b and 36. The chief genera are Cardita
(Jur.-Rec.) and Venericardia (Fig. 24, /, Cret.-Rec.), the

THE LAMELLIBRANCHIA 85

latter being distinguished by the great height and
thickness of the hinge-plate and large size of the teeth.

5. Veneracea. — Concentric ornament, striate or
costate ; occasionally radial also. Valve-margins smooth
or crenulate ; umbones anterior. Sinu-palliate.

The chief genera are Mevetrix (Eoc.-Rec.), already
described ; Venus (Cret.-Rec.), differing from it by its
smaller, pointed pallial sinus, absence of lateral teeth,
and crenulate valve-margins; and Dosinia (Cret.-Rec.),
lenticular, with deep and narrow pallial sinus.

II. Fixed Branch.

Fixation may be temporary or permanent, by a byssus
or by cementation of the shell. The development of a
byssus tends to the abortion of the anterior part of the
body including the anterior adductor, to the development
of ears, and to an inequivalve condition.

I. ARCACEA.

Equivalve;isomyarian; internally porcellanous; hinge-
line with radial teeth, becoming secondarily taxodont;
cardinal area and ligament amphidetic. Ornament radial
or concentric. Fixation never more than temporary.

This sub-order has generally been associated with the
Nuculacea as Taxodonta, but its history shows that the
taxodont structure of the hinge is not primary as in
Nucula, but is derived from a type in which the greater
part of the hinge is occupied by long horizontal teeth, a
few short oblique teeth being present at the anterior end
or in the centre. These characters, as well as the amphi-
detic cardinal area, indicate a relationship to the next
order, Pteriacea, from which the Arcacea diverge in
(a) having lost the nacreous interior, and (b) being almost
always equivalve. Chief genera: Area (Eoc.-Rec.,
with doubtful records back to the Silurian), somewhat

86 PALEONTOLOGY

quadrilateral in shape, with long straight hinge with
vertical teeth and radial ribs (Fig. 25, a); Pectunculus
[Glycimevis], already described (Cret.-Rec.) ; Cucullaa
(Dev.-Rec.), shaped like Area, but with long horizontal
teeth except under umbo, and a raised edge to the posterior
adductor impressions.

II. DYSODONTA.

Usually inequi valve ; hetero- or mono-myarian ; fre-
quently with anterior and posterior " ears," and a right
anterior byssal notch. Cardinal area (if present) and
ligament amphidetic. Fixation temporary or permanent,
by byssus or cementation.

i. Pteriacea. — Inequivalve (except Pevna and Gev-
villia) ; anisomyarian ; internally nacreous ; with ears ;
frequently with a byssal notch below the right anterior
ear ; hinge-line straight, without teeth or with vestiges of
an actinodont type. Silurian to Recent.

A very large and varied sub-order, in which there is a
general tendency (a) to the inequivalve condition, the
left valve being the more convex ; (b) to the lengthening
out of the hinge-line by the formation of " ears " or
" wings " ; (c) to the disappearance of the anterior
adductor. The interior is nacreous, the outer layer
prismatic; cardinal area and ligament amphidetic; hinge
teeth very feebly developed. Ptevia [Avicnla] (Devonian ?
to Recent) differs from Pseudomonotis (already described)
in its more oblique shape and well-marked left anterior
ear (Fig. 25, c, c') ; Pinna (Jur.-Rec.) is acutely triangular
in shape; Conocardium (Dev.-Carb.) is equivalve, inflated,
the anterior ears forming a sort of tube (Fig. 25, I).
These genera are all costate. Perna (Trias.-Rec.) , Gervillia
(Trias.-Eoc.), andlnoceramus (Jur.-Cret.) have the ligament
partly sunk into the hinge-line, and fixed into a long series
of ligament-pits (Fig. 25, d, d')\ these genera are con-

THE LAMELLIBRANCHIA

- lij!liP

& PALEONTOLOGY

centrically ornamented, as a rule : the two first are
equivalve, without curved beaks, while Inoceramus is in-
equivalve, and has beaks incurved. This last attains
great size and thickness in the Upper Cretaceous, just
before extinction, showing the same senile characters as
Hippopodium or some brachiopods — growth of the shell-
margin being quite out of proportion to actual increase
of length. Many other genera might be named.

2. Anomiacea. — Inequivalve; monomyarian ; intern-
ally nacreous; without ears; byssal notch (at least in
early life) converted into a perforation (which may after-
wards close up) ; no hinge-teeth ; ligament amphidetic,
internal.

Anomia (Jur.-Rec.) with the above characters is occa-
sionally found fossil. Placunopsis (Jur.) resembles Anomia,
but the right valve is not perforate and is fixed by
cementation.

3. Pectinacea. — Usually inequivalve ; monomyarian ;
interior lamellar, sub-nacreous ; with thin external
ligament and thicker resilium, both amphidetic ; teeth on
either side of resilium, well-developed or vestigial;
usually with ears, sometimes with a byssal notch. Orna-
ment usually radial. Some genera fixed by byssus or
cementation, others active swimmers. Silurian to Recent.

These shells resemble the Pteriacea in (a), being
usually inequivalve, (b) usually having the hinge
lengthened by " ears," (c) having an amphidetic area and
ligament, and (d) having a byssus, and consequently in
many cases a notch under the right anterior ear. On the
other hand, the ligament tends to sink into the position
of a resilium, between the teeth ; the anterior adductor is
entirely lost, making the shell monomyarian ; and the
internal layer is sub-nacreous. The chief genera are
Chlamys (Trias.-Rec.), both valves rather flat, with good
byssal notch (Fig. 24, e) ; Pecten (Cret.-Rec.), with very

THE LAMKLLIBRANGHIA 89

convex right and flat left valve, and no byssal notch ;
Pterinopecten and Aviculopecten (Sil.-Carb.), with long
hinge and ears not sharply marked off: all these have
feeble hinge-teeth, and are ornamented with radial
costae. In Spondylns (Jur.-Rec.) the cardinal area is
large, especially on the right valve, which is attached ;
the hinge-teeth are large and curved ; the surface is
radially costate and spiny. Plicatula (Trias.-Rec.) is
flatter and without the large area. Lima (Carb. ?-Rec.)
is equivalve, inequilateral, toothless or nearly so, and
with a very feeble byssal notch.

4. Ostracea.— Inequivalve; monomyarian ; interior
lamellar, sometimes sub-nacreous ; hinge-line short,
without teeth, with thick amphidetic ligament ; without
ears or byssal notch ; ornament concentric-striate, or
radial (often very coarse). Fixed by left valve.

These are the oysters. Their very short hinge-line
makes the amphidetic ligament shorter than high and
usually triangular in shape. The subdivision into genera
is unsatisfactory, as the characters that have been relied
upon are such as recur again and again on independent
lines of descent. Thus strongly costate forms have been
named Alectryonia ; forms with the beaks curved spirally
backwards, Exogyra ; those with concave right valve and
large overhanging left umbo, the shell being fixed only
in early life and breaking off by its own weight later,
Gryphaa (Fig. 24, g) ; but these are probably all false or
polyphyletic genera, consisting of a number of species
belonging to different oyster-stocks which have assumed
similar characters by convergence. True genera consist
of species derived from the same immediate stock. Poly-
phyletic genera can be recognized by differences in the
ontogeny of their several species. Gvyphaa in particular
is a series of senile (phylogerontic) forms, the change of
growth-direction being similar to that seen in the huge

90 PAL/EONTOLOGY

forms of Inoceramus of the Upper Cretaceous, or Hippopo-
dium of the Lias.

Ostrea is possibly as old as Carboniferous, and is
abundant from the Jurassic onwards. Gryphaa is mainly
Jurassic; Exogyra, Upper Jurassic and Cretaceous.

5. Mytilacea. — Equivalve; heteromyarian ; interior
nacreous ; umbo anterior ; hinge-line short, without
teeth, or with very obscure cardinal teeth ; ligament
opisthodetic ; no ears or byssal notch, but a slight byssal
gape ; shell elongated in an oblique direction. Ornament
concentric or partly radial. Marine or fluviatile.

This sub-order is typified by the common marine
mussel, Mytilus. The shell is always elongated in an
oblique direction and tends to a triangular shape, the
apex being formed by the umbo, which is at or near the
anterior end of the hinge-line and is never conspicuous,
the base of the triangle being the posterior border. The
area is generally obscure, but apparently amphidetic,
but the ligament is opisthodetic. Hinge-teeth are want-
ing or indistinct. Early genera such as Modiolopsis
(Ord.-Sil.) and Myoconcha (Carb.-Cret.) are less hetero-
myarian than the rest. Modiola (Fig. 25, /, Dev.-Rec.,
maximum in Jurassic) is more quadrilateral than Mytilus,
the beaks not being so far forward. Lithophagus [Litho-
domus] (Carb.-Rec.) is flask-shaped, and bores into corals,
thick shells such as Perna, or rocks on the sea-bottom :
casts of its borings sometimes puzzle the fossil-collector.
Dreissensia (Eoc.— Rec.) and Congeria (Mio.-Plio.) are fresh-
water forms, having the anterior adductor borne on a
myophoric plate. It has been suggested that these are
not true Mytilacea, but homceomorphs derived from some
other stock, but there is little evidence for this.

THE LAMELLTBRANCHTA 91

III. Burrowing Branch.

DESMODONTA.

Usually equivalve, tending to an oblong shape ; often
gaping ; isomyarian ; internally nacreous or porcellanous.
Hinge-line simple, usually one cardinal tooth in each
valve ; ligament opisthodetic. With long siphons ;
usually sinu-palliate. Ornament concentric-striate, rarely
radial.

A number of Palaeozoic genera belong to this branch,
but, owing to imperfect preservation, are not capable of
being further classified : such are Ovthonota (Fig. 25, A,
Sil. ), Grammysia (Sil.-Dev.), and Edmondia (Carb.-
Permian). The Mesozoic and later forms can be more
definitely grouped.

i. Anatinacea. — Equi- or inequivalve ; hinge with
one cardinal tooth in each valve, opisthodetic ligament,
with or without resilium ; internally nacreous (with some
exceptions) ; sinu-palliate.

Pleuvomya (Trias.-Lower Cret.) is elongate-oval, with
fairly strong concentric ornament, and the right valve
slightly overlaps the left along the hinge line. Pholadomya
(Jur.-Rec.) has been already described; it is almost
the only Anatinacean with radial ornament, and is divided
into a number of sub-genera according to shape and
ornament. Thracia (Trias.-Rec.) is inequivalve, the right
valve being the larger; Panopea (Cret.-Rec.) has a thick
shell, gaping at both ends ; Saxicava (Cainozoic and
Recent) is a rock-borer, with the pallial line represented
by a series of disconnected marks. The last two are
non-nacreous.

2. Myacea. — Equi- or inequivalve ; ligament in-
ternal, on a horizontal resiliophore ; internally porcel-
lanous ; sinu-palliate.

92 PALAEONTOLOGY

Corbula (Trias.-Rec.) has already been described. My a
(Eoc.-Rec.) differs in being almost equivalve, and gaping
at both ends ; the anterior adductor is long and narrow,
and the pallial sinus deep.

3. Adesmacea. — Borers in stone or wood, with highly-
modified shell, in which the whole hinge-apparatus
becomes obsolete, while various accessory shelly struc-
tures are added to the valves until finally the latter form
an insignificant part of the skeleton.

Pholas (Jur.-Rec.) has well-developed valves, though
they do not articulate, and is one of the few desmodonts
with radial ornament. The inner layer of the shell is
reflected over the hinge-line and covers the umbo ; there
is a curved rod projecting from the interior of the umbo,
for the attachment of a pedal muscle, and accessory
plates on the outside in the region of the umbo. Teredo,
the "ship-worm " (Jur.-Rec.) bores in floating wood, and
has two small, short valves which are of little service
as a covering to the body, while the immensely long
siphons are encased in a calcareous tube.

Short Bibliography.

MOLL use A GENERALLY.

COSSMANN, M., ET PEYROT, A. — Conchologie Neoge-
nique de 1'Aquitaine, Soc. Linncene Bordeaux (1909-19).

DESHAYES, G. P. — Coquilles Fossiles des Environs de
Paris, 3 vols. (1824-37).

FISCHER, P. — Manuel de Conchyliologie (1880-87).

NEWTON, R. B.— British Oligocene and Eocene Mol-
lusca, Brit. Mus. Cat. (1891).

WOODWARD, B. B. — The Life of the Mollusca (1913).

WOODWARD, S. P.— Manual of the Mollusca (1851-56).
Still very useful, though much out of date in terminology
and classification.

THE LAMELLIBRANCHIA 93

MONOGRAPHS OF THE PAL/EONTOGRAPHICAL SOCIETY,
DEALING ENTIRELY OR PARTLY WITH LAMELLI-
BRANCHIA.

HIND, W. — (i) Carbonicola, Anthracomya and Naiadites
(1894-96). (2) Carboniferous Lamellibranchia, 2 vols-
(1896-1905).

KING, W. — Permian Fossils (1850).

LYCETT, J. — British Trigoniae (1872-79).

MORRIS, J., AND LYCETT, J. — Great Oolite Mollusca,
with Supplement (1850-63).

WHIDBORNE, G. F. — Devonian Fauna (1889-1907).
Cephalopoda and Gastropoda in vols. i. and iii.,
Lamellibranchia in vols. ii. and iii.

WOOD, SEARLES V. — (i) Crag Mollusca, vol. ii.
Bivalves (1851-61). (2) Eocene Bivalves (1861-71).

WOODS, H. — Cretaceous Lamellibranchia, 2 vols.
(1899-1913).

LAMELLIBRANCHIA.

The works listed above are all useful for Lamelli-
branchia. Out of the enormous mass of literature
available, the following are selected as of special value :

BERNARD, F. — Developpement de la Coquille chez les
Lamellibranches, Bull. Soc. Geol. France, ser. 3, vols. xxiii.-
xxv. (1895-97).

DOUVILLE, H. — (i) Classification des Lamellibranches,
Bull. Soc. Geol. France, 1913 ; (2) many papers on Lamelli
branchs, especially the Rudistes, in Bull. Soc. Geol. France
for many years past.

JACKSON, H. T. — Phylogeny of Pelecypoda : Aviculidae
and their allies, Mem. Boston Soc. Nat. Hist., vol. iv., no. 8
(1890).

KITCHIN, F. L. — Jurassic Fauna of Kutch : Lamelli
branchiata, Pal. Indica., ser. 9, vol. iii., pt. 2 (1903).

Ill

THE GASTROPODA

THE Gastropoda, or Snails, are much more active ani-
mals than the Lamellibranchia, most of them crawling
about in search of their food, which they take into the
mouth by means of the radula — a rasping tongue set
with chitinous teeth. The head and foot are perma-
nently outside the mantle-chamber, which is consequently
much reduced in size, with a small opening, and in nearly
all cases lies on one side of the body. This last feature
is part of a profound asymmetry which affects all parts
of the body except the head and foot. In the mantle-
chamber of primitive genera are a pair of gills much like
those of Nucula ; these are reduced to one as a part of
the asymmetry of higher forms, and disappear alto-
gether in the air-breathing forms, the whole mantle-
chamber then serving as a lung-sac. Except in these
last, a ciliary mechanism exists for the purpose of respi-
ration, and in many it also serves to remove the excre-
ment from the mantle-chamber, while in some it even
collects microscopic food (though it cannot convey it to
the mouth, which lies outside the mantle-chamber). The
gastropod shell is distinguished from that of lamelli-
branchs by not being divided into right^ and left valves :

94

THE GASTROPODA 95

it is therefore described as univalve. The simplest
form of this shell is an elliptical cone, the aperture occu-
pying the whole base ; but in the great majority of cases
the shell is a cone coiled in a helicoid spiral.

In many cases there is a horny or calcareous plate
(operculum) which closes the apertures of the shell when
the animal withdraws into it. This, however, is never
hinged to the shell, and the idea that it represents a
second valve like that of lamellibranchs has no justifica-
tion whatever.

Gastropods are the only class of molluscs that inhabit
the dry land, as well as fresh and salt waters. They are
known from all the geological systems,' but it was only in
the Cainozoic era that they became really abundant.
From the Eocene period down to the present day, they
constitute, with lamellibranchs, the main part of the
invertebrate fauna preserved as fossils.

i. Emarginula fissura (Fig. 26), found fossil in the
crags of East Anglia and still living in British seas, is an
example of an almost symmetrical gastropod. It is shaped
like a cone, with an elliptical base and the apex bent over
in the plane of the major axis, which thus appears to be
a plane of symmetry. More careful examination with
a lens shows, especially in youthful examples, that the
apex is curled in a spiral to one side, so that there is no
perfect plane of symmetry. The side on which the
spiral lies is the right side, the direction in which the
apex points being posterior {with reference to the animal's
anatomy). The anterior margin has a deep, parallel-
sided slit. The outer surface is ornamented by alter-

96

PALEONTOLOGY

nalely strong and weak radial ribs, crossed by concentric
ridges (growth-lines), except the apex, which is smooth.
Internally there is seen a horseshoe-shaped scar, marking
the position of the muscles by which the animal pulls
the shell tightly down upon the surface to which it
clings by its broad, muscular foot. The opening of the
horseshoe marks the position of the head.

The notch at the anterior margin indicates the im-
portance of ciliary mechanism in this genus, for it lodges
a process of the mantle containing the exhalent aperture.

FIG. 26. — EMARGINULA FISSURA (LINNE), PLIOCENE. (x2.)

a, Right side view ; a', sma1! portion of surface greatly enlarged to show the
ornamentation which is only diagrammatically shown in the other
figures ; b, view from above ; c, anterior view. Slit, in b and c, black :
at its upper end is seen the callus, and above this the band. (Original.)

Shell-growth in gastropods takes place as in lamelli-
branchs, at the shell-margin and by thickening of the
internal layer : the apex of the shell corresponds to the
umbo. The presence of a marginal notch, however,
causes an interruption to the continuity of the region of
growth. Hence to prevent the notch from becoming
deeper and deeper as growth proceeds it is filled up at
the inner end by a deposit of shell-substance like that of
the inner layer : in time this forms a band from the
apex to the slit that in some species of Emarginula is very
prominent.

THE GASTROPODA 97

This marginal slit, in one form or another, is character-
istic of an important series of gastropods, whose
generally primitive character is confirmed by the fact that
they formed a much larger proportion of the Palaeozoic
than of the modern gastropod fauna.

FlG. 27. — TURRITELLA IMBRICATARIA, LAMARCK, EOCENE.

(Natural size.) (After Deshayes.)

2. Turritella imbricataria (Fig. 27) is a very common
fossil in the Barton beds of the Hampshire coast, and the
Bracklesham beds of the Selsey peninsula. Its shape
approaches that of a very acute cone, but it is formed of
a tube coiled in a spiral, and increasing steadily in size
from the apex of the cone downwards, which is the
direction in which growth takes place. Each turn of
the spiral is called a whovlt and all the whorls except the

7

98 PALEONTOLOGY

last collectively form the spire. Lines drawn at a
tanger.t to opposite sides of the spire will meet beyond
the apex at an angle of 11°, which is the spiral angle of
this particular species. The spiral line which marks the
visible junction of one whorl with the next is called the
suture. The side of each whorl forms an apparently con-
vex curve, but this is seen under a lens to be made
up of eight concave curves, the ends of which correspond
to seven more or less conspicuous ridges which run
spirally round the shell from apex to base, bearing on
them a vast number of fine and close-set tubercles.
Between and parallel to these ridges are much finer lines,
from four to twelve in each interval. These are all
crossed by very fine growth-lines, parallel to the margin
of the aperture.

The aperture, or mouth, of the shell is oval in shape.
When the shell is placed in the usual position chosen for
figures of gastropods— apex upwards* — the mouth is
seen on the right-hand side of the base (this being a
right-handed spiral or dextral shell, as was also Emav-
ginula}. The margin of the aperture is called the
peristome ; it is divided into inner lip (near the middle line)
and outer lip. Living species of Tuvvitella have a horny
operculum, which fits into the aperture when the animal *
is completely withdrawn into the shell ; but this, being
horny, is not found fossil.

If a vertical section is cut through the axis of the
spiral, it is seen that the inner faces of the whorls are

* This is the accepted position in English works, but in French
works it is usual to place the apex downwards.

THE GASTROPODA 99

united into a solid pillar extending from base to apex —

the columella.

The outline of the whorls and detail of the surface-
ornament are specific characters of T. imbricataria .- all
the other characters described (except where common to
all or many gastropods) are generic characters of
Turritella. A shell of this shape, with long spire, and
last whorl not conspicuously larger than the last but one,
is described as turreted. The possession of a solid
columella is denoted by the term imperfoyate ; the simple
outline of the peristome is indicated by the term holosto-
matotts. These terms will be more clearly understood
when the contrary terms are illustrated.

3. Natica multipunctata (Fig. 28) is a common fossil
in the masses of accumulated shells found in East Anglia
and known as Crag, and it still lives at the present day,
though only in warmer seas, as the Mediterranean. The
shape is very unlike that of Turritella, the last whorl
forming the greater part of the external surface, the spire
being very low, and having a spiral angle of 120°. The
whorls are highly convex, nearly semicircular in outline,
but flattened or even slightly concave near the suture.
The large size of the last whorl is not due to an abrupt
increase of sectional area, but to the mode of coiling, the
greater part of each whorl being concealed by the next :
this can be realized either by following in imagination the
extension of the last whorl which would result from
further growth, or by making an actual section through
the vertical axis. Such a section would also show that
instead of being united centrally into a columella the

100

PALAEONTOLOGY

whorls are coiled around a central cavity, the umbilicus,
the presence of which makes this a perforate shell.

The aperture is nearly semicircular, the inner lip being
straight. In the adult animal, a fold of the mantle
extends out over the inner lip, and secretes an- extension
of the inner layer of the shell, which more or less com-
pletely closes over the umbilicus. This deposit is called
callus.

The surface of N. multipunctata is nearly smooth, being

Umb.

-O.L

FIG. 28. — NATICA MULTIPUNCTATA, S. V. WOOD, PLIOCENE.

(Natural size.)
O.L., Outer lip ; Umb., umbilicus. (After S. V. Wood.)

only marked by fine lines of growth and by the puncta-
tions to which it owes its trivial name. In the Crag
specimens the lines of growth are often irregular, some
— often several near together — being more prominent
than the rest. As in lamellibranchs and brachiopods,
such irregularities denote interruptions to the steadiness
of growth, possibly periods of starvation or of a greatly
varying supply of calcareous matter.

Species of the genus Natica are world- wide, frequent-

THE -GASTROPODA 101

ing sandy bottoms in shallow water. In time they range
from at least the Triassic period. They are carnivorous,
using their radula to bore circular holes in lamellibranch
shells, as well as those of other gastropods, through
which they feed on the animal within. Shells thus
bored are very common among the fossils of the Crags,
and of the Miocene of Touraine. Natica, however, is
only one of the genera which bores in this way, and it is
itself sometimes a victim to this mode of attack.

4. Cerithium serratum (Fig. 29) of the famous " Cal-
caire Grossier," the Paris building-stone, of Eocene age,
is a turreted shell, different from Turritella, not only in
its more elaborate ornamentation but in the form of its
aperture, the anterior end of which is produced into a
well-marked channel, the anterior canal. The shell attains
a length of 75 mm. (3 inches). The spiral angle is about
20°. The whorls are flat-sided ; in the first 10 mm. the
ornamentation is weak and then it suddenly becomes
strong. It consists in each whorl of an upper row of
prominent compressed tubercles, 12 or 13 to a whorl,
just below the upper suture ; a row of very fine tubercles,
about 30 to a whorl, three-quarters down the side of the
whorl; and a row of rather larger tubercles, 25 to a
whorl, just above the lower suture. These rows are
crossed by fine lines of growth, which on the last whorl
are seen to have the form of a reversed S, though on the
whorls of the spire the lower limb of the S is concealed.
These growth-lines are parallel to the margin of the
outer lip. The aperture is pear-shaped (the anterior
canal forming the stem of the pear) ; its long axis is at

102 PALAEONTOLOGY

27° to the axis of the spire ; the anterior canal is nearly
straight, but with a slight twist which tends to bring it
more in line with the spire-axis, about two-thirds the
length of the aperture proper, and nearly half its width.
The shell is imperforate. Having an anterior canal it is
siphonostome, Natica and Tnrritella being holostome.

Many species differing in details of ornament from
C. serratum will equally well illustrate^ the genus in the

/ L:

\L

AC,

FIG. 29. — CERITHIUM SERRATUM, LAMARCK, EOCENE, (x^.)

A.C., Anterior canal ; I.L. , inner lip ; O.L., outer lip ; S, suture.
Details omitted near apex. (After Deshayes.)

broad sense. The original genus Cerithium has now been
split up into numerous smaller genera. Recent species
are all marine, and though the group is world-wide, the
most typical species are tropical. The genus is known
fossil from the Jurassic upwards, and had a wide geo-
graphical distribution until at least the Miocene period.

THE GASTROPODA 103

5. Rimella rimosa (Fig- 30) is found in the Barton
beds of the Hampshire coast, and belongs to a family
which is nowadays almost confined to tropical seas. In
shape it differs from turreted shells in that the anterior
part of the last whorl (and of other whorls, though this
is not seen externally) is drawn out to a point of such
length that the last whorl almost repeats the shape of

•P.C

A.C.

FIG. 30.— RIMELLA FISSURELLA (LAMARCK), EOCENE.
(Natural size.)

This species is very close to R. rimosa (Solander). A. C. , Anterior canal :
P.C., posterior canal ; Vt varices. (After Deshayes.)

the spire in inverted position : such a shell is spindle-
shaped or fusiform. The aperture and anterior canal take
part in this drawing-out. The outer lip is thickened, and
the posterior end of the aperture is drawn out into a long
narrow channel (posterior canal) which runs right up the
spire in close contact with it. This canal is analogous
to the notch of Emarginula and its allies, but does not
indicate any near affinity with them.

104 PALEONTOLOGY

The surface-ornament consists of numerous slightly-
curved vertical ridges (corresponding to growth-lines),
the intervals between which are marked by finer spiral
lines, which become broader and flatter as they cross the
vertical ridges.

Careful examination of a well-preserved specimen
shows two things : first, that the actual apex is obtuse,
and the first four whorls are quite smooth ; second, that
at intervals the place of about three of the vertical ridges
is taken by a much broader and more conspicuous eleva-
tion. These elevations, called varices, agree in character
with the thickened outer lip of the aperture ; they are in
fact a series of outer lips successively abandoned as the
shell grew larger. It is evident that a shell with an
elaborate aperture like that of Rimella cannot possibly
maintain the form of its aperture by steady growth,
as Natica or Turritella can do with its simple aperture.
It must either form its aperture once for all and abandon
further growth, or it must alternate between periods of
rapid shell-growth without the specialized aperture, and
periods in which no growth takes place and the elaborate
aperture is in full use. The latter is the case with Rimella :
in place of the occasional irregularities of the growth-lines
of Natica we have periodical growth-periods. What
each of these represents in actual time cannot be stated
with certainty, but probably less than a year.

Another point to be noticed is that none of the varices
of Rimella show any sign of an old posterior canal. This
does not mean that posterior canals were not formed
yntil the adult stage, since immature specimens are found

THE GASTROPODA 105

having a posterior canal: it therefore means that each
posterior canal is actually absorbed during the growth-
period that follows its abandonment. This suggests the
possibility that in some genera the whole of the special
structures at the aperture might be absorbed when
growth restarts. These considerations have a bearing
upon some features found in Cephalopoda (p. 160).

It may be mentioned that in one genus of gastropods,
Distortrix (very rare as a fossil) the method is adopted of
forming an enlarged aperture and afterwards recovering

AC
FIG. 31. — TRIVIA AVELLANA (J. SOWERBY), PLIOCENE, (xf.)

A.C., Anterior canal ; P. C., posterior canal. (Original.)

the normal whorl-form, not abruptly as in other genera,
but gradually : the result is an extraordinary and puzzling
distortion of form.

6. Trivia avellana (Fig. 31), found living in British
seas and fossil in the Crag, is a small hemi-ellipsoidal
shell in which no spiral coiling is recognizable externally,
because the last whorl completely envelops and hides
the spire. Such a shell is said to be convolute. The
aperture is very narrow, nearly parallel-sided and slightly
curved, with slight anterior and posterior canals ; the
outer lip is thickened. The whole surface is ornamented

106 PAL/EON TOLOGY

by ridges disposed more or less at right angles to the
aperture but without the steady parallelism of the spiral
ridges of a turreted shell : these ridges extend over the
edges of both lips, which are therefore said to be toothed.
Trivia is a genus of the family Cypr&idce, the Cowries.
The typical Cypraa is a tropical genus, which includes
the money-cowry and many large and richly coloured
species. It differs from Trivia in having the outer
surface of the shell smooth and polished by the deposit
of an extension of the inner layer by lobes of the mantle
which extend out and lap over it. In the youthful shell
the spire is visible though very short, and only as the
adult stage is reached does the shell become convolute.

7. Fusus porrectus (Fig. 32), of the Barton Clay, is
a many-whorled, fusiform shell. The aperture is pear-
shaped, and is drawn out into a narrow, straight anterior
canal, nearly three times the length of the aperture itself.
The whorls are very convex in outline, ornamented by
nine spiral ridges of unequal strength, crossed by a large
number of varices (12 to a whorl except in some of the
earliest whorls) ; where they cross, the spiral ridges are
particularly prominent. In .the part of the last whorl
corresponding to the anterior canal (a part concealed in
the other whorls) the varices die away and the spiral
ridges are numerous and more uniform. If the shell is
perfect at the apex, the first two whorls are seen to
be quite different from the rest — smooth and globose,
coiled on ah axis inclined at an angle to that of the rest
of the shell. This early stage, so markedly different
from the adult stage, is termed the nucleus or protoconch.

THE GASTROPODA 107

Fust{s is a marine genus, now confined to sub-tropical
seas, but, as in so many other cases, found in Britain in
the Eocene period. It is rare in earlier formations, but
is recorded as far down as the Upper Jurassic.

8. Planorbis discus (Fig. 33) is a common fossil in
the Bembridge Limestone (Lower Oligocene or Sannoi-
sian) of the Isle of Wight. The upper surface is flat
(spiral angle 180°), with a rather deep hollow in the centre.

FlG. 32.— FUSUS PORRECTUS, SOLANDER, EOCENE.

Left-hand figure, natural size (ornament omitted from earlier whorls) ;
right-hand figure, protoconch and post-embiyonic \vhoil. ( x 18.)
(Original.)

The greatest diameter (30 mm. in full-grown shells) can
almost be measured on this upper surface, the margin
being rounded and the outline below quickly turning
inwards at an angle of 35° to the upper surface ; in the
centre of the under surface is a shallow umbilicus, about
one-third the diameter of the shell. The aperture for
two-thirds of its extent has an elliptical outline, but
in contact with the inner whorl this changes to a concave

io8 PAL/EONTOLOGY

curve which meets the general elliptical curve at nearly
a right angle. The shell is thin and fragile, and smooth,
except for faint growth-lines.

The number of whorls is few : on following these care-
fully inwards on the upper surface it is seen that only
the last whorl and a half (in large specimens) has a flat
upper surface ; inwards from thence the surface becomes
convex, and after another turn it plunges downwards
rather rapidly, forming the hollow already noted. The

FIG. 33. — PLANORBIS DISCUS, EDWARDS, OLIGOCENE.
(Natural size.) (After Edwards.)

The lines of growth are too strongly marked in the upper figure.

actual initial part of the shell is thus invisible from the
upper surface, but is plainly visible below, in the centre
of the umbilicus if the latter is free from matrix. It is
clear therefore that the young shell is not only very
different in shape from the adult (more like Natica], but
is a left-handed or sinistral spiral. Indeed it is open to
argument whether the adult shell is not still sinistral, the
flat surface which has been described as upper being
really under. This is the more likely to be the correct

THE GASTROPODA 109

interpretation, since the anatomy of the animal shows it
to be actually sinistral. Other cases are known where
the nucleus or protoconch is sinistral and the adult shell
dextral, and such shells are termed heterostrophic.

Many species of Planorbis are found in the same beds,
and several are also found living in our streams and
ponds, it being a purely freshwater genus. Although
differing in details of shape and proportions from the
species described above, their general likeness to it
is obvious, and in all cases the sinistral character of the
young shell can be recognized.

The classification of the Gastropoda is a very unsatis-
factory matter for the palaeontologist. The features
which zoologists have found to be the best basis for
a natural classification leave no mark upon the shell, and
there are not, as there are in lamellibranchs, important
shell-characters that can be made a basis for broad
divisions. If gastropods were entirely or mainly extinct,
we should be driven to make a shell-classification, but it
would be quite obviously artificial. As it is, there are
few extinct genera which are not obviously related to
recent forms, so we have to accept the zoological classi-
fication and place these extinct genera with their nearest
allies.

The fundamental division is based upon a difference in
the nervous system : in one main sub-class, the Strepto-
neura, the twisting of the body which brings the anus
round nearly to the head has twisted the nerve-loop which
supplies the gills into a figure-of-8 ; in the other, the
Euthyneura, it has left it straight. The latter sub-class

i io PALEONTOLOGY

falls naturally into two divisions, the Opisthobranchiata,
marine forms with gills, and the Pulmonata, terrestrial
and freshwater forms in which the mantle-chamber is
converted into a lung-sac. These two latter groups are
so distinct that many systematists raise them to the rank
of sub-classes, and as this is convenient for palaeonto-
logical purposes we will do so here. No shell character
can be pointed to as distinctive of either of these three
sub-classes ; nevertheless, any one who is familiar with
gastropod shells generally would feel little hesitation in
placing a genus he had never seen before in its proper
sub-class. He would not find it easy to state his reasons,
as they would consist in combinations of characters,
positive and negative, which could only be stated with
many qualifications. Among other general facts that he
would have in mind would be the much larger proportion
of fossil genera belonging to the Streptoneura than to
the Euthyneura (for in the latter, and especially in the
opisthobranchs, there is a tendency towards the reduction
and final disappearance of the shell) ; and the restriction
to freshwater deposits of Pulmonata and a few genera of
Streptoneura.

SUB-CLASS: STREPTONEURA.
ORDER: ASPIDOBRANCHA. '

SUB-ORDER : Docoglossa. — This includes the limpets,
forms with shells like Emarginula, but without the notch.
Forms undistinguishable from the common modern
limpet, Patella, date from the Silurian, and allied recent
genera are also known as fossils.

THE GASTROPODA in

SUB-ORDER : Rhipidoglossa. — This includes all the
remaining nacreous gastropods with others that are
porcellanous. Palaeontologically they may be divided
into two series, according to the (a) presence or (b)
absence of the lateral slit (described in Emarginula.}

(a) Fissttridea resembles Emarginula in shape, but the
notch becomes closed in at an early age by the union
of the margin and forms a " key-hole " in front of the
apex ; as the shell grows this perforation becomes larger,
by resorption of the shell, until it occupies the whorl
apex and is far from the margin. In Haliotis, the "ormer "
of the Channel Islands or "Venus' ear," new notches
are formed as the earlier are first closed around and
eventually filled in, so that at any time one notch and
several perforations are in use. The important fossil genus
Pleurotomaria(Fig. 34, a, Sil.-Rec. ) is spirally coiled and has
a slit, the filling in of which gives rise to a spiral band,
towards which the lines of growth are indented in a
V-like manner. The geological history of Pleurotomaria
is similar to that of Pholadoniya or Trigonia among
lamellibranchs : world-wide in its distribution up to the
Cretaceous period, it then became restricted and now
survives only in the seas of Japan and the East and
West Indies.

Bellerophon (Fig. 34, b, Sil.-Perm.) is one of the few
gastropod shells which is coiled symmetrically, so that it
resembles externally a cephalopod shell (but has no
internal septa) : the last whorl more or less completely
envelops the others, leaving either a very narrow umbilicus
on each side, or none. Euomphalus (Fig. 34, c, Sil.-Trias.)
is a discoidal shell — that is, one in which the spiral angle
is increased to 180° or even more, so as to become a re-
entrant.

(b) Among genera without a lateral notch, the most
important are Trochus (Sil.-Rec.), a conical shell with

112 PALEONTOLOGY

a flat base and an aperture wider than high (Fig. 34, d) ;
Turbo (Sil.-Rec.) differing from Trochus in having a convex
base and a nearly circular aperture, so that if placed apex
downwards it is seen to resemble a spinning-top : this
form is described as turbinate ; Nerita (Jur.-Rec.), globose,
imperforate with semicircular aperture and thick, often
grooved, lips ; Neritina (Eoc.-Rec.)j similar, but with thin
shell and lips, a fluviatile snail (Fig. 34, e).

ORDER: CTENOBRANCHIA.

Divided into a number of sub-orders on a basis of
tongue-structure, the only convenient palaeontological
division is into holostome and siphonostome forms.

(a) Holostome. — Scalaria (Trias.-Rec.) is turreted,
with circular aperture, strong, sharp vertical ridges
and finer spiral lines. Solarium (Jur.-Rec.) is conical,
with obtuse spiral angle, umbilicus very wide and deep,
and four-sided aperture. The nucleus (or larval shell) is
heterostrophic, as in PlanovUs, as can easily be detected on
a well-preserved specimen.

Littorina (Jur.-Rec.), the periwinkle, somewhat re-
sembles Turbo in form, but is not nacreous. Natica and
Turritella have already been described. Vivipam, or
Paludina (Jur.-Rec.), is turbinate with rounded apex,
rounded whorls and a sub-circular aperture : it is a fresh-
water form (Fig. 34, g).

The family Capulida is an interesting case of de-
generation : it has taken to microscopic food and a
fixed habit ; in consequence the shell reverts to a more
or less limpet-like form. There is therefore some un-
certainty as to whether certain fossils belong to this
family or to the Patellidce. Capulus (Fig. 34, /) shows
a trace of spiral curvature at the apex ; Crepidula
(Cret.-Rec.), the slipper-limpet, has a slipper-shaped
shell, and forms arch-like colonies, one individual on

THE GASTROPODA 113

top of another ; Calyptraa (Cret.-Rec.j, the bonnet-
limpet, or Chinaman's hat, is depressed-conical, with
a spiral shelf in the interior ; Platyceras (Camb.-Trias),
limpet-like, with spirally-rolled apex, often found adhering
to the ventral surface of crinoids, on which it was in
some degree parasitic. The Capididce attained their
acme in the Silurian and Lower Devonian, particularly
of Bohemia, where many strange forms occur, including
some which are high-conical and almost cylindrical.

The family Melaniidte, with the turreted genera Melania
(Jur.-Rec.) and Melanopsis (Cret.-Rec.), is transitional
to the siphonostome group, the latter genus having an
anterior canal and the former not. Melania (Fig. 34, h)
is distinguished from Turritella by the shape of the whorls,
which gave a step-like outline to the shell, and the
greater strength of the vertical ornament, which develops
strong tubercles or spines.

(b) Siphonostome. — Cerithium has been described
already. Closely allied to it is Potamides, which at present
lives in mangrove swamps and the estuaries of tropical
and sub-tropical rivers. There is no absolute way of dis-
tinguishing this genus from Cerithium by shell-characters
alone, though the opercula are distinct, and in the fresh
shell there is a dark green periostracum (external non-
calcareous layer of shell, rarely preserved in fossils).
Consequently the reference of fossil species to Potamides
is partly based upon their occurrence along with un-
questionable fresh-water shells. Potamides is divided now
into several sub-genera : it is known from the Cretaceous
upwards, and had a much wider distribution in the
Eocene and Oligocene periods than now, being common
in the strata of the Isle of Wight. These two genera
are members of the family Cerithiidtt, of which other
members occur as far back as the Triassic period. In
these early genera the anterior canal is very feebly

8

H4

PALEONTOLOGY

J K.

FIG. 34. — GASTROPODS.

a, Pleurotomaria mosensis Buvignier, Upper Jurassic. (Xi.) (After
Buvignier). L.S., Lateral slit, b, b' . Bellerophon hiulcus J. de C.
Sowerby, Lower Carboniferous. (x£.) (After de Koninck.) c, c' ,
Euomphalus pentan%ulatus J. Sowerby, Lower Carboniferous. ( X |.)
(After Goldfuss.) d, Trochus inornatus Buvignier, Upper Jurassic.

THE GASTROPODA 115

marked, so that its development must have taken place
gradually within this family.

In the Mesozoic rocks there is found another family,
the Nerinaida, some members of which have great super-
ficial resemblance to Centhium in form and ornament, so
that they can only be distinguished by taking vertical
sections through the shell (Fig. 34, i). It is then seen
that the cavity of each whorl is more or less constricted by
internal spiral ridges (not confined to the columella as any
are that occur in Centhium}. In extreme cases these
ridges are so developed that only a narrow space of com-
plicated form is left for the snail's body ; for instance,
in Ptygmatis, a common genus in the English Lower
Oolites.

Rimella belongs to the family Strombida, other genera
of which have the outer lip expanded and sometimes
branching in a finger-like manner. The family ranges
from the Jurassic period to the Recent, and though
abundantly represented in the British area up to and in
the Eocene period, it is now confined to the Indo-Pacific
province, the Mediterranean and West Indies. The
closely-allied family Apovrhaida, however, with similar
digitations to the outer lip (Fig. 34, j}, and having the
same range in time, is essentially Atlantic in its modern
distribution, and the species Aporrhais pes-pelicani is
common in the British seas.

Trivia and the Cypraida have already been mentioned.

FIG 34. — GASTROPODS (continued}.

( X 3. ) (After Buvignier.) e, Nerita aperta J. de C. Sowerby, Oligo-
cene. ( X \. ) (After Edwards.) /, Capulus ungaricus (Linne") , Pliocene.
(Xj.) (After S. V. Wood.) g> Vivipara lenta (Solander), Oligocene.
(Xf. ) (After Deshayes. ) h, Melanatria inquinata (Defrance),
Eocene. (Xf.) (After Deshayes. ) i, Nerincea moreana d'Orbigny,
Upper Jurassic. (Xj.) (After Buvignier.) j, Aporrhais parkinsoni
Mantell, Gault. (Xj.) (After Starkie Gardner.) k, Nassa reticosa
(J. Sowerby), Pliocene. (X |.) (After S. V. Wood.) In several cases
the ornamentation is omitted from the earlier whorls for the sake of
clearness of outline.

ii6 PALAEONTOLOGY

Allied to them is Pymla (Cret -Rec.), another of the
forms, now exclusively tropical, found in the British
Eocene : a pear-shaped form with very short, obtuse
spire, and very large aperture with thin outer lip.

Another group of siphonostomes is represented by the
whelks, Buccinum and Nassa (Tertiary and Recent), the
former of boreal and the latter of general distribution.
These are elongate-oval imperforate shells with spires of
moderate length, wide apertures and short anterior canal,
ornamented by coarser vertical and finer spiral lines.
In Buccinum the canal is in line with the long axis of the
aperture ; in Nassa (Fig. 34, k) it makes a considerable
angle with it. The latter genus is also distinguished by
crenulations on the interior of the outer lip. Another
allied genus is Neptunea [Chrysodomus], a boreal form,
nearly smooth, notable for its left-handed species which
abound in the Red Crag (Fig. 35, a)

Murex (Cret. -Rec.) differs from the whelks, firstly in
having a very long and narrow canal, the edges of which
bend over until they all but meet (split-tube structure^,
and a very rounded aperture ; secondly, in the thicken-
ing of the mouth-border, from which and along the
anterior canal there often arise long spines which repeat
the split-tube structure. These mouth-borders persist as
varices, often of great regularity in arrangement, there
being exactly three to a whorl, so that those of successive
whorls continue one another in a vertical (or very slightly
oblique) line up the spire ; it has been stated that in
these forms one turn of the spire represents a year's
growth. In other cases the varices are more numerous.
In Typhis (Eoc.-Rec.) the anterior canal and spines
become perfectly tubular. Murex is now widely dis-
tributed except in the cold seas, Typhis is more closely
confined to warm waters, but, like so many others, is
found in the British Eocene.

THE GASTROPODA 117

Fusus has been described; allied to it is Clavella
(Fig. 35, &), an abundant Eocene form, now represented
by a single Polynesian species. It differs from Fusus in
being smooth and in the shape of the whorls, which have
rather flat sides terminated above by a horizontal shelf;
Sycum, or Leiostoma (Eocene), is another smooth form,
differing in its much shorter spire and less clearly
denned canal.

The Volutida (Cret.-Rec.) consists of somewhat fusi-
form shells, though without the sharp demarcation of the
anterior canal seen in Fusus, the aperture being long and
somewhat parallel-sided. Their special feature is the
development of spiral plaits or ridges on the columella,
which define grooves along which slide the tendons by
which the animal pulls itself into the shell. The com-
monest fossil genus is Volutilithes (Cret.-Rec.), a richly
ornamented form, abundant in the British Eocene
(Fig. 35, c), but now represented by a single South African
species. Aurinia, or Scaphella (Fig. 35, e), a smooth form
with very rounded apex, is found in the British Pliocene,
and now lives in warmer seas.

Another fusiform shell with aperture shaped as in the
Volutida is Pleurotoma (Cret.-Rec.), in which there is a
posterior canal which takes the form of a notch at a little
distance from the suture (Fig. 35, d), recalling the lateral
slit of Pleurotomayia, and like that causing an inflection of
the growth-lines, but not closed by a special secreted band.
This abounds in species in the British and other Tertiaries,
and in the warm seas to-day. Cotiorbis (Eocene) is similar,
but with narrower aperture, forming a transition to Conus
(Cret.-Rec.), with very short, depressed spire, narrow,
parallel-sided aperture, and posterior canal close to the
suture : its distributional history is like that of Pleurotoma,
but it is more closely restricted to tropical seas to-day.

Brief mention must be made here of a specialized

PALEONTOLOGY

FIG. 35.— GASTROPODS.

Neptunea contraria (Linne"), Pliocene. (x£.) (After Harmer. )
b, Clavella longteva (Lamark), Eocene. (X^.) (After Deshayes.)
A.C., Anterior canal; P.C., posterior canal, c, Volutilithes athleta

THE GASTROPODA i*9

group of ctenobranchs, the Heteropoda, whose shells are
very thin and transparent and symmetrical in their
coiling. Both these changes from typical gastropod
characters indicate adaptation to a pelagic life — that is,
one of swimming in the open sea. Transparency is the
most effective form of camouflage for a pelagic animal
which has nowhere to take cover, and the lop-sided shell,
which does very well for a bottom-crawler, would be a
serious impediment to active swimming. Heteropoda
are very abundant in the Mediterranean and other warm
seas ; but fossil examples are rare. The symmetrical
character of the shell of Bellerophon led to its being at one
time placed among the heteropods, but its shell is much
thicker, and the presence of £ slit indicates its rhipido-
glossal affinities. In the Gault occurs a little shell of
similar form, Belleropkina, which would have better claim
to be considered a heteropod, were it not that it had
a nacreous inner layer.

SUB-CLASS: OPISTHOBRANCHIA.

The Opisthobranchia are divided into two sub-
orders, Tectibranchia and Nudibranchia. The latter are
shell-less and unknown fossil. The former is divided
into a number of families, but on a basis of shell-form
they may be very roughly grouped into (i) more or less
turbinate forms, such as Action, distinguished from
Streptoneura of similar shape by the strongly- marked
folds on the columella, and (2) forms such as Butt a and
Scaphander (Fig. 35, /), with a more or less completely con-

FIG. 35. — GASTROPODS (continued).

(J. Sowerby), Eocene. (Xj.) (After Deshayes). d, Pleurotoma
transvenaria Lamarck, Eocene. (X$.) (After Deshayes.) e, Sca-
phella lamberti (J. Sowerby), Pliocene. (X|.) (After S. V. Wood.)
/, Scaphander edwardsi (J. de C. Sowerby), Eocene. (x£.) (After
Dixon.) g, Hyolithes princeps Billings, Lower Cambrian. (X§.)
f, Side view ; g1 , back view ; g" , operculum. (After Walcott.)
A, Limncea fusiformis J. Sowerby, Oligocene. (X f.) (After Edwards. )

120 PAL/ttONTOLOGY

volute shell, distinguished from Cypvcea, etc., by the shape
of the aperiure (narrow posteriorly, widened anteriorly).
There is also a pelagic group known as the Pteropoda,
occupying a similar position among opisthobranchs to
that of the Heteropoda among Streptoneura, their shells
having become superficially symmetrical, and very thin
and translucent, by adaptation to a swimming life in the
open sea.

The Pteropoda, according to the investigations of
Prof. Pelseneer of Ghent, do not form a natural group,
being composed of two series derived from separate
families of opisthobranchs. Nevertheless the name is a
convenient one and may well be used if it is understood
that it has not an exact taxonomic value. Pteropods
abound in all the modern oceans to such an extent that
where there is moderately shallow water far from land
their shells form an important proportion of the forami-
niferal ooze on the sea-floor — so-called pteropod ooze (at
greater depths their thin shells dissolve before reaching
the bottom). In the Cainozoic rocks of the Mediter-
ranean region, similar pteropod-bearing deposits are
known, and some pteropod shells have been described
from the Cretaceous ; but the entire absence of ptero-
pods from the pelagic Chalk of Europe, where they
might particularly be expected, seems to show that this
specialized group was then at the very beginning of its
existence ; and it is not surprising that the Jurassic
and Triassic rocks contain no certain remains of ptero-
pods. It is therefore startling to find in the Devonian
rocks abundant shells indistinguishable from those of the
modern pteropod Styliola ; and from thence downwards
to the Cambrian there are found a number of genera
which, though less closely similar to modern forms, are
yet more like pteropods than anything else (Hyolithes,
Fig. 35,g, Conularia, Tentaculites, etc.). Recently, among the

THE GASTROPODA 121

wonderfully preserved fossils from the Middle Cambrian
of British Columbia, Mr. Walcott has found a Hyolithes
showing swimming organs closely resembling those of
modern pteropods. It is impossible that pteropods can
have existed before the opisthobranchs from which they
are derived, and no opisthobranch is known before the
Carboniferous period. It is also very unlikely that any
group should have abounded in the Palaeozoic era, have
lain hidden through the Mesozoic, and have again
become abundant in the Cainozoic. The most probable
explanation is that we have here again a case of homceo-
morphy, due perhaps to adaptation to like conditions.
These Palaeozoic forms should therefore be removed
from the Pteropoda and placed provisionally in a separate
order for which the name Conularida may be used. It
is quite likely that this is not a natural group : some of
its members may possibly have been tube-secreting
annelids.

SUB-CLASS: PULMONATA.

These are terrestrial or freshwater forms, with shells
of very variable shape, but always thin and unorna-
mented (except by colour, which is sometimes preserved
in fossils). Planorbis has been described; other fresh-water
forms are Lzww^fl, with long pointed few-whorled spire and
oval aperture (Fig. 35, h), and Physa,oi somewhat similar
shape, but sinistral. Among land-snails are Helix, the
common snail, Pupa, which has a cylindrical form, the
maximum diameter being attained very early, and
Glandina, resembling Limnaa in form.

All these genera are common in the numerous fresh-
water beds of the Tertiary. Previous to that freshwater
deposits are scanty. The oldest, in the Devonian, have
yielded no gastropods, though they contain freshwater
lamellibranchs ; and in the Coal Measure no freshwater

122 PALEONTOLOGY

pulmonates are found, but there are a few terrestrial
forms, such as Dendropupa and Archaoxonitts, genera
almost identical with the Pupa and Zonites (allied to
Helix) of to-day. The oldest known freshwater pul-
monates appear to be Physa and Planorbis of the
Purbeck beds.

Short Bibliography.

(See under " Mollusca Generally," p. 92.)

GRABAU, A. W. — (i) Studies of Gastropoda, American
Naturalist, vols. xxxvi., xxxvii., xli. (1902-07); (2) Phy-
logeny of Fusus, Smithsonian Misc. Coll., vol. xliv. (1904);
(3) the above are summarized in Grabau's Principles of
Stratigraphy, p. 967 et seq.

MONOGRAPHS OF THE PAL^ONTOGRAPHICAL SOCIETY.

BLAKE, J. F. — Fauna of the Cornbrash (1905 07).

HARMER, F. W. — Pliocene Mollusca, vol. i. (1913-19)

HUDLESTON, W. H. — Gastropoda of the Inferior
Oolite (1887-96).

MORRIS, J., AND LYCETT, J. — Great Oolite Mollusca,
with Supplement (1850-63).

SLATER, I. L. — British Conulariae (1907).

WHIDBORNE, G. F. — Devonian Fauna (1889-1907).
Gastropoda in vols. i. and iii.

IV
THE CEPHALOPODA

THE present-day Cephalopoda (cuttle-fish, squid, octopus
and nautilus) are the most highly organized of Mollusca.
They agree with gastropods in having a rasping tongue,
but the crawling foot is replaced by a series of powerful
arms around the mouth, and the mantle and body-wall
become muscular swimming organs. While the crawling
life of gastropods has made them asymmetrical, and only
a few have acquired a superficial symmetry through
taking to a swimming life, nearly all cephalopods retain
the perfect bilateral symmetry, while as active voracious
animals their sense-organs have become highly de-
veloped. Only a few extinct forms are asymmetric, and
these may well have been crawlers exclusively, while
the rest of those known as fossil may, like the modern
members of the class, have been at least capable of
swimming and sometimes swimmers exclusively. The
habitat of cephalopods is more restricted than that of
gastropods, being exclusively marine.

Most of the modern cephalopods, such as the cuttle-
fish and squid, show no sign of a shell externally,
though it is to be found in a modified condition buried in
the interior of the body ; some, as the octopus, have

123

124 PALAEONTOLOGY

none at all. Only three living genera have an external
shell: that of Argonauta, the "paper nautilus," is of
a special type peculiar to it, and need not be considered
here ; that of Spirula is almost embedded in the mantle,
but is of the same general type as the completely external
shell of Nautilus (the " pearly nautilus "). This last is a
spirally-coiled univalve shell, divided internally into
chambers, and the same general type is found in an
enormous number of extinct Cephalopoda. As the
spiral coiling introduces a complication into the structure,
it will be convenient to describe first the similar but
straight shell of the extinct genus Orthoceras, filling up
the gaps in our knowledge by reference to the living
nautilus. There are a great many species of Orthoceras,
ranging in age from uppermost Cambrian (Tremadocian)
to Triassic, but as none of these is so common as to be
obtainable with certainty for examination, and it may
often be necessary to take several species to demonstrate
the full structure, the description here given is generic
instead of specific.

i. Orthoceras. — The shell (Fig. 36) is a cone ap-
proaching to a cylinder (sometimes circular, sometimes
elliptical in plan), and ending in a more or less blunt
rounded apex. In extremely rare cases there is found
attached to this apex what looks like a sraall shrivelled-
up bladder, and this is probably the horny equivalent of a
globular calcareous body in Spirula — the protoconch, or
initial shell, secreted in early life. The main shell (or
conch as it is sometimes called) is divided internally into
chambers by partitions (septa), which are convex towards

THE CEPHALOPODA

125

the protoconch and are perforated in (or near) the centre
by an opening around which the septum extends like
a bottle-neck (Fig. 36, S.N.). At the opposite end from
the protoconch a considerable portion of the cavity of the

FIG. 36.— ORTHOCERAS.

a, a', a", O. capax Barrande, Silurian, a, Cast of body-chamber and
three gas-chambers, part of test remaining (shown by fine striations) ;
a', median section of a similar broken specimen ; a", view of end
broken along a septum, b, O. decipiens Barrande, Silurian, cast
showing perfect body-chamber with constriction, septa more closely set
than in O. capax. c, O. capillosum Barrande, Silurian, median section
showing perfect body-chamber without constriction, gas-chambers
partly filled, and continuous siphuncle. (All after Barrande.) B.C.,
Body-chamber; C, constriction; G.C., gas-chamber; 5, septum;
Siph., siphuncle ; S.N., septal neck ; Sut., septal suture ; T, test.

shell is undivided, forming the body-chamber, lodging
the animal's body. The rest are gas-chambers, con-
taining (in Nautilus] a gas secreted from the blood,
resembling atmospheric air but with more nitrogen.

126 PAL/EONTOLOGY

Through the central perforations of the septa ran a
living cord, the siphuncle, through all the gas chambers
back to the protoconch. While the animal was growing
new shell continually was being added to the aperture or
" mouth " of the body-chamber; while at intervals the
animal moved forwards in its body-chamber, and secreted
behind it a new septum. Thus any given septum was
formed at a rather later time than the part of the external
shell with which it is in contact — a fact that must never
be forgotten. The line of junction of the edge of a
septum with the external shell js termed the suture (more
precisely the septal suture, for in coiled cephalopods the
term suture is also used in the same sense as in gastropods,
for the line of external contact of the whorls). * The
septal sutures are invisible externally, but when the shell
is removed the internal cast shows them.

The ornamentation of Orthoceras is usually of a very
simple character; consisting of growth- lines parallel to
the margin of the aperture — usually only fine striae, at
other times with stronger rings at intervals. The actual
aperture of the body-chamber is rarely preserved (the
thin shell, unsupported by septa, being easily crushed) ;
but in a few perfect specimens the aperture is seen to be
a simple circle, which may or may not have just behind
it a circular constriction (Fig. 36, b).

2. Nautilus radiatus is a fairly common species in
the Lower Chalk, and the following description applies
to it, but, with slight alterations, would serve for many
other species. It is a shell symmetrically coiled in
a plane-spiral, so tightly that each whorl nearly conceals

THE CEPHALOPODA

127

the whorl within, leaving only a very deep and narrow
umbilicus on each side. The shape of the whorl-section
thus comes to be that shown in Fig. 37 (right-hand
figure), and the outline may be divided into a peripheral
area, a pair of lateral areas, a pair of umbilical zones,
and an impressed area, of which the last is concave,
the rest more or less convex (see Fig. 49, e, for these

FIG. 37. — NAUTILUS ALBENSIS, D'ORBIGNY, CENOMANIAN, DERIVED
(CAMBRIDGE GREENSAND).

Internal cast. Body-chamber at least missing. Umbilicus filled with
phosphatic matrix. Side view and front view. (Natural size.)
(Original. )

terms). The peripheral area is known from the anatomy
of living species to be ventral, so that the umbilical and
impressed areas are dorsal. The ornamentation consists
of a series of slight corrugations which are essentially
growth-lines, and correspond to the aperture-margin, but
owing to the spiral coiling there has to be more rapid
growth on the ventral than on the dorsal side, and the
growth-lines are not parallel as they practically are in

1 28 PALEONTOLOGY

gastropods, but radiating : in cephalopods they are there-
fore usually termed the radial lines. These lines are very
faint near the umbilicus, but become more distinct in the
outer part of the lateral areas, where they swing back-
wards and cross the peripheral area in a curve which is
concave forwards (towards the aperture). This is called
the hyponomic sinus (cf. Fig. 45, a), and corresponds to an
embayment of the aperture occupied by the funnel or

Siph u n cle

FIG. 38.— NAUTILUS EXCAVATUS, J. DE C. SOWERBY, INFERIOR
OOLITE.

Approximate median section (not accurately median at the centre).
(Natural size.) (Original.)

hyponome, an organ by which the nautilus ejects water
from the mantle-chamber and so projects itself through
the sea. This hyponomic sinus is characteristic of most
Nautiloidea, though wanting in the primitive species of
Orthoceras. If the fossil is broken, the septa are seen to
be concave forwards as in Orthoceras, with the siphuncle
placed rather nearer the dorsal side (Fig. 37, right-hand
figure). On the cast the sutures are seen to run radially

THE CEPHALOPODA i2<)

outwards from the umbilicus for half the height of the
lateral area, then to swing forwards, and to cross the
peripheral area with a very slight backward curve : they
thus contrast sharply with the radial lines, of which they
cross four or more in their course.

3. Asteroceras obtusum (Fig. 39) is the index-fossil
of one of the zones of the Lower Lias. Specimens of this
or closely allied species are familiar to visitors to Lyme
Regis. The shell is of a light yellow-brown colour, but
is usually only preserved in fragments, the fossils being
essentially internal casts. The shell is spirally coiled in
a plane— or bilaterally — symmetrical spiral. Each com-
plete turn of the spiral is termed a whorl. The number
of whorls cannot usually be counted, the central being
hidden by rock-matrix which is difficult to remove, but
usually four or more can be counted : they become larger
rather rapidly from the centre outwards. The body-
chamber occupies rather less than half the last whorl : it
is filled with a grey matrix of argillaceous limestone,
while the gas-chambers are filled with brown crystalline
calcite. The visible part of the shell encircled by the
last whorl is called the umbilicus. Unlike that of gastro-
pods this forms a concave area on both sides equally.
The spiral line of contact between each whorl and the
next is the whorl-suture (corresponding to the suture of
a gastropod, but present on both sides of the am-
monite).

A plane-spiral shell is measured as follows : the
diameter is measured from the outer edge of the aperture
(or if the shell is imperfect, of the latest part preserved)

9

"130 PALEONTOLOGY

straight across the centre to the opposite edge. The
measurement of a single whorl in the same direction
gives its height; at right angles to this is its thickness.
In an ammonite with strong ribs, such as Asteroceras,
this thickness will vary very much according as it is
measured at or between the ribs. The width of the
umbilicus is measured from whorl-suture to whorl-
suture along the same line as the diameter. These are
the most important dimensions, and it is their relative
proportions, rather than their absolute amounts, that are
of value for comparison of one species with another. As,
however, these proportions frequently alter in the process
of growth, the actual size of an ammonite measured must
always be stated. The four essentials of ammonite
measurement are, then — (i) diameter; (2) height of last
whorl; (3) thickness of last whorl; (4) width of umbilicus.
The first may be stated in millimetres, the others as
percentages of the first. Thus for a particular specimen
of Astevoceras obtusum we express these dimensions thus :

H5> 36> 27'3I> 41-

The 27-31 refers to the difference of thickness between
and at the ribs.

For comparison we give measurements of two allied
species, A steroccras redcarense :

118-5: 36, 30-34, 37-5,
and Astcroceras olifex :

88 : 34*5, 25-27> 38-
A glance at the three formulae shows us that A. redcarense

THE CEPHALOPODA

FlG, 39. — AsTEROCERAS OBTUSUM, J. SOWERBY, SlNEMURIAN
(LOWER LIAS), LYME REGIS.

a, Side view; 6t apertural view (Xj.), (Original); c, suture-line (after
d'Orbigny. The arrow denotes the middle line of the periphery and
points towards the body-chamber ; the horizontal line is the guide-line,
EL, External lobe ; ES, external saddle; LLi, LL2, first and second
lateral lobes; LSi, LSz, first and second lateral saddles; AL, AS,
auxiliary lobe and saddle.

132 PALAEONTOLOGY

is thicker, A . olifex thinner than A . obtusuw, while both
are less widely umbilicate.

If we place the specimen so that the aperture faces us
(apertural view, Fig. 39, b) we see the shape of the whorl
in cross-section. It is somewhat elliptical, and the outline
may be divided into six parts, corresponding to six regions
on the surface of the whorl : (i) the somewhat flattened
peripheral or external area or venter, which bears a broad
rounded keel in its middle line; (2) and (3) the lateral areas,
one on each side, slightly convex but with a general
direction approximately parallel to the plane of symmetry;
(4) and (5) the inner or umbilical areas ; one on each side,
which slope inwards to the whorl-sutures ; and (6) the
concave impressed area, invisible except when the whorl is
detached from the shell, lying between the two whorl-
sutures and in close contact with the next inner whorl.
The ratio between the height of this overlapped part of
an inner whorl and the height of the outer whorl which
overlaps it is the amount of indentation of the outer whorl ;
while the ratio between it and the total height of the
whorl of which it forms part is the amount of inclusion
of the inner whorl. In A. obtusum the former is about \,
the latter about | (the height of the outer whorl being
about double that of the next inner). These ratios are
low when the umbilicus is wide (as in Asteroceras], but
they increase as it becomes narrower.

The spiral line along which the peripheral area meets
the lateral area is called the peripheral margin ; that on
which the lateral and inner areas meet is the umbilical
margin (as for descriptive purposes the umbilicus is taken

THE CEPHALOPODA 133

to include the inner area of the outer whorl, but not for
purposes of measurement). In Asteroceras these margins
are not sharply defined, as they are in some ammonites.

The ornament of Asteroceras consists of a series of
strong ribs, which start at the umbilical margin and
run almost straight outwards until near the peripheral
margin, where they bend decidedly forwards and die
away. Between their ends and the median keel, the
periphery forms a shallow groove on either side. The
forward bend of the ribs shows that the aperture instead
of having a hyponomic sinus, as in Nautilus, has a
median projection or rostrum (cf. Fig. 45, /).

The ribs are well spaced- out, their distance apart being
proportionate to the height of the whorl ; thus in the last
whorl of the specimen figured it gradually increases from
8 mm. to 1 6 mm. (as measured in the middle of the
lateral area). Careful examination of the inner whorls
shows that the ribs there are not so smooth as in the
outer whorls : each is thickened to form a tubercle about
the middle of the lateral area. This ammonite, as it grew
up, suffered a degeneration (catagenesis) of its ornament,
from more to less complex.

The ribs and tubercles are not thickenings of the shell,
but corrugations, and therefore show as well on the cast
as when the shell (or test) is present.

The septal sutures (Fig. 39, c) are very different from
those of Orthoceras or Nautilus : they are thrown into a
series of forward and backward curves, each of which is
further puckered up or frilled. The forward curves
(convex towards the body-chamber) are called saddles, the

134 PALEONTOLOGY

backward curves are lobes. The visible suture-line (without
the portion on the impressed area) consists of (i) a
median peripheral or external lobe, divided into two by a
small pointed saddle, (2) a pair of external saddles on the
peripheral margins, followed on each side by (3) the first
or superior lateral lobe, (4) the first or superior lateral saddle,
(5) the second or inferior lateral lobe, (6) the second or inferior
lateral saddle, (7) an auxiliary lobe, (8) an auxiliary saddle,
(9) another auxiliary lobe. Both lobes and saddles are
much subdivided, but the numerous subdivisions are all
approximately equal. The first lateral saddle is much
larger than the second, and lies but little above the
centre of the lateral area.

If we draw a straight line from the centre of the
ammonite to the ends of the external lobe (normal line or
guide-line), we see that all the other lobes fall short of this
line, the first lateral lobe most of all.

4. Xipheroceras planicosta is another ammonite
from the zone of A. obtusutn. It is common at Lyme
Regis, but far more beautiful specimens are, or were,
found in a bed of limestone at Marston Magna near
Yeovil — the " Ammonite Marble " or Marston Marble,
at one time much used for ornamental purposes, and still
generally obtainable in small fragments from mineral
dealers. In both localities the usual specimens are
small (diameter about 2 cm.), larger ones (5 or 6 cm.)
are rare and probably not identical in species. The
dimensions of one of the small size are

20: 33, 32'36»45-

THE CEPHALOPODA 135

Whorl - section reniform (kidney- shaped) in inner
whorls, becoming, higher and less broad later. Amount
of indentation one-tenth, of inclusion one-sixth (at 20 mm.
diameter). When the umbilicus is clean there can be
seen in the centre a small globular protoconch, with six
whorls around it (up to 20 mm. diameter). The first
two whorls or so are quite smooth, then fine striae
appear; after another whorl and a half obscure ribs are
seen, and about the end of the fourth whorl well defined
ribs appear. At the 20 mm. diameter the ribs are about
2 mm. apart, and run as follows : starting in the steep
inner area they first slope backwards, then curve round
to a truly radial direction on the umbilical margin, from
which they curve very gently to a slight backward
slope on the outer half of the lateral area ; at the peri-
pheral margin they swing forward and cross the periphery
without interruption in a forwardly-convex curve, but at
the same time they become much broader and lower (as
though modelled in a plastic substance and then pressed
down). The fine striae on and between the ribs follow
the same direction, which shows that here as in Asteroceras
there was no hyponomic sinus to the aperture but a
rostrum, though not a very prominent one. Larger
specimens of this (or nearly allied) species show that
soon after attaining 20 mm. diameter the ribs begin
to show signs of tubercles at the peripheral margin, and
in some the flattened peripheral rib tends to divide into a
bundle of three ribs. In another species (X. ziphus) the
tubercles are developed into long spines. Thus the early
species of Xipheroceras show an increasing strength and

136 PALEONTOLOGY

complexity or anagenesis of ornament with advancing age,
in contrast to the catagenesis of ornament in Asteroceras.
It also shows decided change in whorl-shape.

The body-chamber of these small specimens occupies
just over half a whorl, and the last two or three septa
are usually crowded together — a feature common in the
last septa of large ammonites. This suggests that the
small size of these ammonites was not due to premature
death, but that they were a small species. Larger
species of the same genus are for the most part rather
later in time, so that the genus shows anagenesis in
size as well as in ornament.

The suture-line of Xipheroceras is (at 20 mm. or less)
like a much-simplified Asteroceras suture, except that there
are no auxiliary lobes and saddles, and the lateral lobes
do not lie so much in front of the peripheral. In older
shells the subdivision of the lobes and saddles becomes
much more complex.

These four examples of cephalopods have been chosen
for description as easily obtainable, so that readers can
verify the descriptions on actual specimens. The next
few examples are less common, and will be described
more briefly.

5. Tetrameroceras obovatum (Fig. 40) is one of
several more or less ellipsoidal species found in the Lower
Ludlow shales of Herefordshire and Shropshire. It is
about no mm. long, 58 mm. wide, and 75 mm. high.
Preserved as a cast, it is seen that about one-half of the
shell is body-chamber, the rest consisting of about nine
gas-chambers. The whole shows a very slight curvature,

THE CEPHALOPODA

137

the side which we shall presently recognize as ventral
being slightly concave : such a curvature is the opposite
to that of Nautilus, where the ventral side is convex.
This is expressed by saying that Tetrameroceras is endo-
gastric, Nautilus exogastric. A more striking difference
from both Orthoceras and Nautilus is the form of the
aperture, which is greatly narrowed in the centre, forming
a slit only 3 mm. wide. This widens at the ventral end

Br.

v

a

FIG. 40.— TETRAMEROCERAS OBOVATUM (BLAKE), SILURIAN
(LOWER LUDLOW BEDS).

a, Side view; b, apertural view, (xg.) (After Blake.) Br., Brachial
portion of aperture; H.S., hyponomic sinus; D, dorsal side;
V, ventral side.

into an oval, at the dorsal end into a four-lobed expansion
Evidently the animal was permanently shut in its body-
chamber, and could only protrude two pairs of arms for
obtaining food by the larger four-lobed aperture, and the
swimming funnel by the other. The latter therefore
represents the hyponomic sinus and fixes the side
towards which it lies as ventral. In other species con-
temporary with this there may be openings for only

138 PALEONTOLOGY

one pair of arms (genus Gomphocems) or for three or four
pairs.

6. Meloceras [Cyrtoceras] elongatum (Fig. 41, a)
is found in the Upper Silurian limestone of Bohemia. It
is compressed-elliptical in section and curved in form, the
curvature decreasing from youth to age (i.e., from apex
to body-chamber), and the plane of curvature being that
of the major-axis of the ellipse. It attains a length of
80 to 90 mm. of which nearly a third is body-chamber.
The septa have the simple form seen in Orthoceras, but
are more closely set than in that genus, their distance
apart being only about one-sixth the major axis of the
cross-section. The siphuncle is close to the convex
margin, and forms a continuous calcareous tube, con-
stricted as it passes through the septa. The shell is only
ornamented by fine striae, the course of which shows
a very slight hyponomic sinus on the convex margin,
which is therefore ventral. Thus this species is exo-
gastric, and the siphuncle is ventral (its usual position
when not central). There are other species attributed to
the same genus which are endogastric.

7. Cophinoceras [Gyroceras] ornatum is found in
the Middle Devonian limestone of the Eifel. Compared
with the last it has a stronger curvature, so that it forms
a complete spiral of at least a whorl and a half, but there
is no contact between the whorls. It is depressed-
elliptical in section (i.e., the long axis of the ellipse is at
right angles to the plane of coiling). The only ornament
seen on the cast is a paired series of rather coarse and

THE CEPHALOPODA 139

ill-defined tubercles along the peripheral margins, but on
the shell itself the periphery is marked by longitudinal
ridges and transverse striae. The septa are farther apart
than in Melocevas ; the siphuncle in the same position.

8. Ophidioceras simplex (Fig. 45, b), from the
Silurian of Bohemia, is a flat-sided spiral shell, 35 mm.
in diameter, of several whorls which are just in contact,
except that a very small portion of the last whorl becomes
straight and so separates from the next inner whorl. The
aperture is constricted and Y-shaped, the hyponomic sinus
(the stem of the Y) being towards the periphery, so that
the shell is exogastric (as are most, if not all, of the spiral
cephalopods). The sides are ornamented by simple ribs,
interrupted by a keel on the periphery.

These three examples illustrate stages in the transition
from a straight shell like Orthoceras to a tightly coiled
one like Nautilus. At one time these different degrees of
curvature were regarded as markingMifferent genera, and
generic names were given accordingly. Now it is recog-
nized that they are evolutionary stages which may have
been passed through by different stocks at the same or at
different times. Names for these stages are useful.
Following Mr. Buckman in modifying the more cumbrous
terms of Hyatt, we may call the straight shell an
orthocone ; the curved one which does not form a spiral.
cyrtocone ; the open spiral, a gyrocone ; the spiral with
whorl-contact, an ophiocone. The terminology does not
at present distinguish between evolute ophiocones like
Ophidioceras, in which the whorls merely touch, and those
in which, by tighter coiling, the inner whorls become

140 PALEONTOLOGY

more or less included by the outer. By tighter and
tighter coiling (a continuation of the same process which
led from cyrtocone to ophiocone) the umbilicus becomes
smaller and smaller until we reach the sphcerocone with no
umbilicus or a very small one, as in Nautilus.

If these stages represent a real evolutionary series, we
ought to find (i) that they follow one another in time,
and (2) that the ontogeny of forms in the later stages
should show traces of the earlier stages.

As to the first point, remembering the imperfection of
the palaeontological record, and the fact that different
stocks must have gone through similar developments at
different times, we must not expect too much. Orthocones
are known from the Lower Cambrian to the Trias, and
are pretty equally distributed though getting scarcer after
the Carboniferous. Cyrtocones are known from the upper-
most Cambrian, but are most abundant in the Silurian and
Devonian ; gyrocones, rare in the Silurian, are abundant
in the Devonian, and then become rarer ; ophiocones,
though known from the Ordovician and Silurian, are rare
until the Devonian and Carboniferous; while sphaerocones
first appear in the Carboniferous and continue to the
present. Thus, on the whole, the known sequence in
time agrees with the expected sequence.

As to the second point, allowing for the tendency to
hurry through ancestral stages during ontogeny, the
evidence is satisfactory. The theoretically earlier stages
do often precede the later in ontogeny, as may be seen in
several illustrations (Figs. 41, b\ 45, a}.

At the same time, ontogeny shows evidence of

FIG. 41.— CEPHALOPOD COILING CYCLE.

a, a

Mcloceras [Cyrloceras] elongatum Barrande, Silurian. (X^.)
Cyrtocone : a, side view, striate ornament indicated at one place,
part of test removed, showing suture-lines and part of siphuncle (s) ;
a', cross-section, b, Inner whorls of Agoniatites fecundus (Barrande),
Middle Devonian (X^), showing orthocone, cyrtocone, and ophiocone
stages ; P, protoconch. c, Spiroceras calloviense (Morris), Upper
Jurassic (Kellaways Rock), (X2); c' , initial part (Xio); P, proto-
conch. This shows uncoiling from serpenticone to criocone. d, Mor-
phoceras dimorphum (d'Orbigny), Bajocian. Ornament omitted.
Shows uncoiling, sphserocene to serpenticone ; also five constrictions.
e, Turrilites bechei Sharpe, Cenomanian. Ornament omitted except on
last whorl, and one suture-line indicated. /", Toxoceras bituberculatum
d'Orbigny, Neocomian. Toxpcone. g, Scaphites cequalis J. Sowerby,
Cenomanian. Scaphiticone, beginning as sphaerocone and ending
in a hook, h, Ancyloceras pulcherrimum d'Orbigny, Barremian.
Criocone changing to toxocone, with hook-like end. Ornament
omitted except near aperture, i, Macroscaphites ivani (Puzos),
Barremian. (Xj.) Serpenticone uncoiling, with hook-like end. Orna-
ment omitted except on body chamber. j, Hamites, Albian. A
generalized figure, a, b, c, After Barrande ; e, after Sharpe ; the rest
after d'Orbigny. All x j, unless otherwise stated.

M2 PALEONTOLOGY

another tendency — that of reversion in old age to earlier
stages. Thus Meloceras elongatum tends to an orthocone
stage by lessening of curvature ; Ophidioceras reverts
abruptly to the same. In Palaeozoic times examples of
this are not very frequent, but among Mesozoic. Cephalo-
poda a tendency to uncoil is common : the next few
examples will illustrate this.

Spiroceras calloviense (Fig. 41, c, c') is a common
fossil in the Kellaways Rock of Wiltshire. At first sight
it appears to be a gyrocone, but close examination shows
that it begins with a globular protoconch, around which
the tube-like shell winds closely for more than one
whorl, after which its curvature diminishes and it con-
tinues as an open (evolute) spiral. As this is a stage in
an uncoiling process it must not be called a gyrocone : the
term applied is criocone.

Toxoceras (Fig. 41, /) is a genus of which various
species are known from Cretaceous rocks. It has the
form of a cyrtocone, and the protoconch and initial part
of the conch never seem to be preserved, so that direct
evidence of uncoiling is wanting ; but as it occurs so late
in time, and has sutures like those of contemporary coiled
and uncoiling forms, it is reasonable to regard it as a
stage in uncoiling (toxocone}.

Baculites (Fig. 46, /, I'} is a long, straight, compressed
form, found in the uppermost Cretaceous strata. Its
initial stages are usually lost, but in specimens from
South Dakota they have been found to be substantially
similar to those of Spiroceras, the closely-coiled stage

THE CEPHALOPODA 143

lasting rather longer. It is thus the final stage of
uncoiling (baculicone}.

These three forms illustrate the reversal of the process
which led from orthocone to ophiocone. There are many
indications that the closely-coiled forms which precede
them in the Jurassic and Cretaceous are themselves
derived from sphaerocones, so they may be termed
serpenticones. Thus we have a double series, one of
increasing curvature (anagenesis] and one of decreasing
curvature (catagenesis}.

Sphaerocone

./ Ni

Ophiocone Serpenticone

/" \

Gyrocone Criocone

Cyrtocone Toxocone

/• \

Baculi

Orthocone Baculicone

Such a double series forms an evolutionary cycle. It
should never be possible to confuse the corresponding
stages of the upward and downward series, because

(1) the ontogeny of a perfect specimen should show some
trace of the stages earlier in the series, and if this fails

(2) there are other features, such as the sutures, which
have their own independent anagenesis and catagenesis,
and their stages will never exactly correspond with the
coiling stages. Thus we find degenerate suture-lines in
Cretaceous serpenticones, but not in criocones, etc. This
is an example of the principle of iyreversibility in evo-
lution, according to which an organism which reverts to

M4 P

the condition of one of its remoter ancestors never does
so entirely : it retains features acquired by its inter-
mediate ancestors. It should thus always be possible to
distinguish degenerate from primitive forms.

As with the anagenetic series, so with the catagenetic :
the same stages are passed through by different stocks at
different times. Uncoiling forms, including a baculicone,
occur as early as the Noric (Upper Triassic), where
they are contemporaneous with the last orthocones ; but
it is not until the Cretaceous that they become abundant.

Just as the early ophiocones often show a tendency to
revert to the straight form in old age, so the uncoiling
forms show a tendency to coil again in old age, producing
hook-like body-chambers. Ancyloceras ends thus after a
criocone stage, Macroscaphites after a serpenticone stage,
while Scaphites passes direct from sphaerocone to this
hook. Hamites (initial stages unknown) makes two
hook-like bends. All these are Cretaceous (Fig. 41, g-j),
but in the Jurassic period CEcoptychius had the form of
Scaphites (but different ornament), and there were a
number of sphserocones which showed a tendency to
uncoil in old age (scaphitoids). Sometimes there are
signs in ontogeny of uncoiling followed by re-coiling.

All these forms so far described are bilaterally sym-
metrical, but there are a few asymmetrically-coiled
forms (turricones), resembling sinistral turreted gastropods,
mainly Cretaceous (e.g., Tnrrilites, Fig. 41, e) but with one
example in the Noric; and one extraordinary form
(Nipponites) which can best be described as a three-
dimensional zigzag. By analogy with the gastropods,

THE CEPHALOPODA 145

which are mostly asymmetric and crawlers, but in
which the few swimming or floating forms become
symmetric, we may suppose these few asymmetric
cephalopods to have been crawlers (benthic), while the
rest were, at the least, not confined to the bottom, even
if they were not habitually swimmers (nektic) or floaters
(planktic).

The thin shell with its large volume of gas-chambers
must have been a very light object in the animal's life-
time, enabling the animal to rise or sink in the sea with
ease, as does the modern Nautilus. After death it might
float and be carried by currents to a considerable distance.
Thus the three surviving species of Nautilus live, in the
Polynesian seas between Sumatra and Fiji, but their
empty shells are found on the coast of Japan and else-
where; the three species of Spirula live in deep sub-
tropical waters, and only about half a dozen specimens of
the animal have ever been found, but the empty shells are
found in great numbers on the coasts of New Zealand,
and are also known in the Banda Sea, the Canary Islands,
West Indies, etc., and have even been found on the
Cornish coast. Thus it may be the case with fossil
cephalopods that the geographical distribution of the
shells is far wider than that of the living animals. Mr.
Buckman points out, however, that in travelling long
distances such delicate shells could not escape injury;
so that when we find (as we frequently do) fossil cepha-
lopods with the most delicate structures preserved
uninjured, it is reasonable to suppose that they lived
where they are now found ; but when we find the shells

10

146 PALEONTOLOGY

showing signs of travel — broken body-chamber, abraded
surface, or general damage, they may fairly be regarded
as having floated from a distance.

In the process of burial on the sea-floor, the sediment
would fill the body-chamber, and penetrate through the
septal neck into the first gas-chamber, which it could
only fill up to the level of the entry. A little might
penetrate into the second chamber, but beyond that it
would have great difficulty in making its way unless
the shell was broken in places (though fine calcareous
mud is sometimes found in the air-chambers of apparently
unbroken shells). The empty chambers would rarely
remain, empty, however : water would permeate them
and deposit materials from solution. Thus in calcareous
rocks the chambers are usually filled or partly filled with
crystalline calcite ; in clays, with pyrite or marcasite.
Where very little sediment was being deposited, they
were often filled with calcium phosphate. As a conse-
quence, fossil cephalopods, especially those of large size,
are usually very heavy objects, and it needs an effort of
imagination to realize them as possibly floating animals.
Further, the thin shell itself may have been removed, in
one of two ways : by solution in the rock, or by flaking
away during extraction, so that we have an internal cast.
When the ornament is a corrugation, not a thickening,
of the shell, it shows as well on the cast as on the shell,
but the casts are at once distinguished by their showing
the septal sutures, which of course are on the inner,
not the outer, surface of the shell.

Sometimes the shell itself may be replaced by silica or

THE CEPHALOPODA 147

pyrites : in which case no sutures are visible on the
surface. Only the finest ornament in cephalopods is
usually a thickening (capillae) or thinning (striae) of the
shell, but there are cases among ammonites where pro-
tuberant parts of the surface — whether corrugations
(costae) tubercles or keel (carina) — have their cavities
cut off from the general shell-cavity by a partition, so
that the cast fails to show these elevations or shows
them in a much smaller degree than the shell-surface.
Mr. Buckman terms these tubular structures septicostce
(found in dactyloids), septitiibenles (in deroceratids), and
septicarin& (in many oxycones). The partitions are
wanting in the greater part of the body-chamber, being
found in its posterior part and in the gas-chambers, so
that they are secreted by the hinder part of the mantle,
and like the chamber-septa are later in date than the
adjacent shell. These structures may be of considerable
stratigraphical value — e.g., serving to distinguish Upper
Lias dactyloids from Upper Jurassic perisphinctids,
even in fragments.

Among existing Cephalopoda two orders may be re-
cognized, distinguished by the number of gills and various
other features. The order Tetrabranchiata (four-gilled)
includes the single genus Nautilus; all other living
forms belong to the Dibranchiata (two-gilled). It has
been usual to extend this classification to extinct forms,
but while many of these are so evidently allied to the
recent forms that they can safely be assigned to one or
other of these orders, there is one very important series
of fossils which diverges quite as much from the nauti-
loids in some respects as do the dibranchs. It is there-

148 PALEONTOLOGY

fore advisable to recognize three orders, and to one of
these no name having reference to gills can be applied
since gills are not preserved in the fossils.

The three orders may be briefly characterized thus,
but it must be understood that as the two other orders
diverge from the ancestral order Nautiloidea, there are
transitional forms which may not conform strictly to the
definitions.

1. The Nautiloidea are the ancestral and most con-
servative order. Commencing as orthocones at the end of
the Cambrian period, they pass through the intermediate
stages to the sphaerocone in the Palaeozoic era, but sub-
sequently remain without further change down to the
present time. Only a very few forms, and those quite
early, show any tendency to uncoil.

Generally distinctive features are — (i ) the protoconch is
not calcareous and is nearly always shed ; (2) the septal
sutures are either simple circles or they become slightly
wavy ; they are never frilled ; (3) the siphuncle is usually
near the centre of each septum, very rarely is it in contact
with either margin ; (4) the septal necks point away
from the body-margin ; (5) fhere is a hyponomic sinus.

2. The Ammonoidea contrast with the steady-going
Nautiloidea: they had a short life and a merry one.
Diverging from primitive Nautiloidea at the end of the
Silurian epoch, they hurry through the early stages of
coiling, and in the Mesozoic era begin to uncoil. Some
of them uncoil completely as early as the late Triassic
period ; others at various stages in the Jurassic ; others
after some uncoiling start coiling-in again ; but in the
Cretaceous period uncoiling becomes very general and is
followed by the abrupt extinction of the order (though
coiled forms also persist to the last).

General distinctive features are more difficult to state
than in the case of the Nautiloidea, because of the great

THE CEPHALOPODA 149

changes through which the order passes, but the following
may be given :

(i) The protoconch is calcareous and is never shed;
(2) the septa are like those of nautiloids in the most
primitive families, but soon become folded and eventually
frilled at the edges, so that the suture-lines attain a very
high degree of complexity ; (3) the siphuncle migrates to
one of the margins, in the great majority to the periphery ;
(4) the backwardly directed septal necks (retro sip Jionate]
soon become replaced by forwardly projecting septal collars
(prosiphonate] ; (5) the hyponomic sinus is present in most
Palaeozoic genera, but afterwards disappears, and may
be replaced by a rostrum.

3. The Dibranchiata vary so much in shell-structure
that it is difficult to make any statement that shall apply
to the whole order. The main feature is the general
subordination of the chambered shell (phragmocone] to
other skeletal structures, resulting in a great delay of
the tendency to coil. The septa are simple, the siphuncle
more or less marginal and retrosiphonate.

Among the large number of Palaeozoic nautiloids we
may choose a few for mention.

(1) Genera retaining the straight form of Orthoceras,
but distinguished from it by peculiar features of the
siphuncle. In Endoceras (Ord.-Sil.) the siphuncle is nearly
half the width of the shell and the septal necks are
funnel-like and fit in to one another, forming a continuous
tube. In Actinoceras the siphuncle is central, and swells
out in each gas-chamber, but is constricted as it passes
through the septa. Internal casts of the siphuncle of a
gigantic species of this type resemble a row of vertebrae
(Huronia vertebralis) when weathered out from the Ordo-
vician limestone of Drummond Island, Lake Huron.

(2) Orthocones or cyrtocones which acquire peculiar

150 PAL/KONTOLOGY

features when full grown. Tetramerocevas (p. 136) is an
instance; Gomphocems (Ord.-Sil.) is an exogastric form
with T-shaped aperture; Phragnwceras (Sil.) an endogastric
gyrocone. In Ascoceras (Sil.), after growing as an ortho-
cone to some length, the shell suddenly expanded, the
earlier portion was detached, and the few septa that were
afterwards formed are so strangely shaped that the gas-
chambers are on the dorsal instead of the posterior side
of the animal.

(3) Ophiocones (Fig. 45, a) are found in the Devonian
and are abundant in the Carboniferous. Sphaerocones
begin with Solenocheilus of the Carboniferous, and Nautilus
(Trias. -Rec.) continues this form. Hercoglossa (Trias. -
Cainozoic) and Aturia (Cainozoic) are remarkable as
mimicking the Devonian clymenids (see below), both
in the folding of their septa and the dorsal position of
the siphuncle.

(4) Uncoiling forms are very rare among nautiloids,
and are confined to early times. Lituites is Ordovician,
Ophidioceras (Fig. 45, b) is Silurian : both are ophiocones
(sometimes beginning as cyrtocones) which revert to
orthocone form, the former at an early period of its life,
the latter only in old age. Otherwise the Nautiloidea
seem to have found stability and permanence in the
sphserocone stage.

The Ammonoidea are more interesting to the geologist
than are the Nautiloidea, because of their great value
as zone-fossils, which arises from the wonderful vitality
of the stock, which during six geological periods threw
off swarm after swarm, each going through an evolu-
tionary cycle with rapidity and then either dying out
or giving rise to a new stock repeating a similar cycle,
until at last exhaustion is reached in the Cretaceous
period and the whole order dies out.

THE CEPHALOPODA 151

Examples of some of the most highly organized
ammonoids (Asteroceras, Xipheroceras] , as well as of
degenerate forms (Spiroceras, Toxocems, Baculites] have
already been described. Examples of. the most primi-
tive forms will now be given.

In the Upper Devonian shales of Biidesheim in the
Eifel district of Western Germany, there are found great
numbers of internal casts of small tightly-coiled cephal-
opod shells, in pyrites oxidized on the surface to a rich
brown colour. They rarely exceed 15 mm. in diameter,
but they are practically complete shells, since they show
the cast of the body-chamber as well as the septate
portion. They vary very much in relative thickness
and degree of involution, so that a number of different
species are distinguishable, but if the septal sutures are
examined they can at once be sorted into two series.
We choose an example of each to describe.

4. Manticoceras [Gephyroceras] retrorsum has (in
the specimen shown in Fig. 42) a diameter of 15 mm.

and its proportions are —

»

15: 53> 57> 20.

The body-chamber, so far as preserved, is not quite
half the last whorl in length. The umbilical margin falls
very steeply, almost perpendicularly near the aperture,
but less steeply in the earlier-formed part of the last
whorl. The lateral and peripheral areas seem at first
sight to pass insensibly into one another, but very careful
examination reveals a very faint longitudinal depression
about 2^ mm. on either side of the middle line : this is

152 PALEONTOLOGY

more pronounced in the earlier formed part of the last
whorl. It may be taken as marking the peripheral margin.
The ornamentation, as seen in the cast, consists of two
quite independent series of lines — (i) a series of very
faint ribs, most clearly seen at the peripheral margin,
where they are less than i mm. apart. They are curved
with a forward concavity on the lateral area, and at the
peripheral margin they bend backwards and cross the
periphery in a deep curve like the hyponomic sinus

FIG. 42.— MANTICOCERAS AFF. RETRORSUM (VON BUCK), UPPER
DEVONIAN; BUDESHEIM, EIFEL. (x2.)

Cast of an immature example ; the faint ornamentation omitted ; urn-
bilicus filled with matrix ; numerous suture-lines behind the body-
chamber. (Original.)

of Nautiloidea. (2) The other ornament is a series of
much finer and closer irregularly-sinuous lines, about
five to each of the ribs, to which they are not parallel, but
run obliquely back from the umbilical margin in a curve
which is at first convex forwards and is afterwards
reversed.

The suture-line (septal suture) is of a kind intermediate
between those of a Nautilus and an ammonite : lobes and
saddles are well-defined, but entirely free from frilling
(goniatitic type). There is a broad peripheral lobe,

THE CEPHALOPODA 153

divided into two by a small median saddle ; a wide
external saddle on each side, and a rounded lateral lobe
(Fig. 47, b). It is the special feature of the family
Gephyroceratida that the external lobe and saddle are so
wide that the latter comes to lie in the middle of the
lateral area ; hence many authors call it the " lateral
saddle." In ontogeny the little median saddle appears
late, so that there can be no doubt that the divided lobe

FIG. 43.— ^MANTICOCERAS RHYNCHOSTOMA J. M. CLARKE, UPPER
DEVONIAN ; BIG SISTER CREEK, ERIE COUNTY, NEW YORK,
U.S.A. (x£.)

Adult, with rostrum and lappets. Last suture-line only shown.
(After Clarke.)

of which it forms part is the external lobe and not a pair
of lobes.

Although the Eifel specimens are never much larger
than the specimen described, in beds of the same age in
New York there have been found specimens of a closely-
allied species which grew to a great size : in these the
hyponomic sinus disappeared in age and was replaced
by a forward projection or rostrum (Fig. 43).

154 PALEONTOLOGY

5. Tornoceras simplex is said to attain a diameter
of 75 mm., but it is not often that specimens are found
much larger than that figured, the size and proportions
of which stated in the method already explained are

I4'5-55»47'5» -•

The umbilicus is either entirely closed, or too small to be
measured accurately.

The body-chamber extends through about half a
whorl (it may be longer). The cast is nearly smooth,
but shows growth-lines (radial lines) whose course is
similar to that of Manticocevas. The suture-line, how-
ever, is quite different : the external lobe is narrow and
undivided ; on each side of it is a rounded external
saddle, a rounded lateral lobe and a rounded lateral
saddle (Fig. 47, c). It is the lobe that comes in the
middle of the lateral area, so that the contrast with
Manticoceras is striking.

6. Gastrioceras carbonarium (Fig. 44) occurs in
some abundance in the marine bands which form the
" roofs" of certain coal-seams in the Lower Coal Measures
of Lancashire and Yorkshire. Specimens have been
found as much as 160 mm. in diameter, but a more usual
size is about 30 mm. and one of these small specimens
gives the following proportions (the query to the last
item denotes difficulty of measurements owing to the
umbilicus being filled with tough shaley matrix) :

30 : 45-5, 60, 40 ?
The inner area falls very steeply from the umbilical

THE CEPHALOPODA 155

margin. The periphery is arched and cannot be marked
off from the lateral area. The greater part of the
surface is marked by very fine striae, which on the lateral
area are straight and radial but cross the periphery
in a forwardly-concave curve, indicating a much
shallower hyponomic sinus than in the Devonian forms
already described. The striae are not uniform, but are
arranged in bundles, which in passing from the periphery
towards the umbilicus take on the character of ribs and

FIG. 44. — GASTRIOCERAS CARBONARIUM (VON BUCH), WESTPHALIAN

(COAL MEASURES), LANCASHIRE.

(Natural size.)

Immature ; two constrictions shown ; the striate ornament indicates a
hyponomic sinus (right-hand figure). (Original.)

at the umbilical margin project slightly as blunt tubercles.
At intervals of about half a whorl there occur " constric-
tions " — i.e., depressions of the shell, i mm. or more in
width, following exactly the course of the ornament-
lines.

These constrictions are analogous to the varices of
gastropod shells : each one was originally formed just
behind the apertural margin (as in Orthoceras decipiens,
though in that species there is only one), and it may be
presumed that its formation was followed by a period of

156

PALEONTOLOGY

FIG. 45.— APERTURAL MARGINS OF CEPHALOPODA.

a, Discitoceras leveillianum (de Koninck), Carboniferous Limestone,
Clane, Ireland, a, Side view, showing gyrocone stage in centre,
passing into ophiocone ; there are indications of the radial and longi-

THE CEPHALOPODA 157

cessation of growth, after which renewed growth went
on rapidly for half a whorl when another constriction was
formed.

Most specimens preserve the shell, but even where it
is chipped away no suture-lines can be seen. This is
because the body-chamber was over a whorl in length :
only on the inner whorls of broken specimens can the
suture-line be seen (Fig. 47, d). It resembles that of
Tornoceras in having a lateral lobe and saddle, but the
lobe is pointed instead of rounded ; while the external
is divided by a median saddle which is much more
prominent than that of Manticoceras retrovsum.

These three Palaeozoic species are conveniently
described as "goniatites," being characterized by suture-
lines with simple lobes and saddles.

FIG. 45. — APERTURAL MARGINS OF CEPHALOPODA (continued)

tudinal ornament, a', Peripheral view, showing hyponomic sinus
(H.S.) (After Foord and Crick.) b, Ophidioceras simplex Barrande,
Silurian, Lochkow, Bohemia, b, Side view, showing ophiocone finally un-
coiling, and aperture with hyponomic sinus (H.S. ). b' ', Peripheral view ;
b", aperiural view. (After Barrande.) c, Arcestes distincttts(G\e\&\),
Noric (Alpine Trias). Sphaerocone ; no distinct hyponomic sinus,
rostrum, or lappets. ( After Quenstedt.) d,d' \Protrachyceras archdaus
(Laube), Ladinic (Alpine Trias). Apertural region, showing broad
rostrum. (After Mojsisovics. e, Paltopleuroccras pseudocostatum
(Hyatt), Domerian (Middle Lias). Long crenulate rostrum. (After
Quenstedt.) /,/"', Pleydellia snbcompta (Branco), Yeovilian. Showing
relation of radial line to apertural margin. (After Buckman.)
g, g' , Reineckia exstincta (Quenstedt), Vesulian. Showing tendency
to constriction of aperture, absence of distinct rostrum and lappets,
and relation of radial line to apertural margin. (After Quenstedt.)
h, h' , Brasilia effricata S. Buckman, Aalenian, Bradford Abbas, Dorset.
Showing very broad rostrum, no definite lappets, and relation of radial
line to apertural margin. (After Buckman.) (All X J.) H.S., Hypo-
nomic sinus ; L, lappet ; ft, rostrum.

158

PALEONTOLOGY

FIG. 46.— APERTURAL MARGINS OF CEPHALOPODA.

a, a', Phylloceras mediterraneum Neumayr, Vesulian, near Digne (Basses
Alpes). (x£.) Showing rostrum and lappets (unusual in Phyllocera-
tidce] and a number of varices, only parallel to apertural margin if

THE CEPHALOPODA 159

GENERAL CHARACTERS OF AMMONOIDS.

The body-chamber of the typical ammonoids varies
from half a whorl to over two whorls in length. The
aperture of nearly all Devonian forms has a hyponomic
sinus on the periphery, but in the Carboniferous period
some of the goniatites lose this, and begin to develop a
slight forward projection (rostrum) in its place. In the
Triassic and early Jurassic periods this rostrum becomes
more and more prominent, until in some cases it projects
as a narrow rod with a length greater than the height of
the whorl (Fig. 45, c-f). Sometimes the rostrum may

FIG. 46. — APERTURAL MARGINS OF CEPHALOPODA (continued}.

rostrum and lappets are excluded. (After Haug.) b, Hildoceras sp. ,
Whitbian (Upper Lias). Rostrum, lappet, and radial line. (Alter
Wright.) c, Lucya cavata S. Buckman, Aalenian (concava zone),
Bradford Abbas (Dorset). Constriction behind aperture. (After
Buckman.) d, Oppelia fusca (Quenstedt), Vesulian, CEschingen
(Wiirttemberg. j Lappet constricted at base ; relation of radial line to
aperiural margin well shown. (After Quenstedt.) e, e' , Fontannesia
cuivata S. Buckman, Bajocian (discites zone), Bradford Abb.is
(Dorset). Very ieeble rostrum, constricted (spoon-like) lappets.
(After Buckman.) /,/', Witchellia deltafalcata ((Quenstedt), Bajocian.
Rostrum almost lost, lappets bent inwards to constrict aperture.
(After Quenstedt.) g, Cosmoceras elizabethce (Pratt), Callovian (Lower
Oxford Clay), Christian Malford (Wilts). Long lappets, no rostrum ;
relation of radial line to apertural margin. (After Quenstedt.) h, h' ,
Otoites sp., Bajocian, Lauffen (Wiirttemberg). Incurved lappets, no
rostrum ; no evidence of lappets in radial line. (After Quenstedt.)
i, i' , Morphoceras pseudo-anceps (Ebray), Vesulian, St. Honore-les-
bains (Nievre, France). (x§. ) Lappets and bifurcated rostrum
uniting to divide aperture into five orifices — A, median hyponomic
orifice ; B, paired opening, probably for eyes ; C, paired opening for
arms. (After H. Douville".) /, /, Siemiradzkia comptoni (Pratt),
Callovian. Long incurved lappets, no rostrum, a constriction shown
some distance behind aperture ; radial line parallel to constriction,
not to apertural margin. (After Quenstedt.) £, &', Sphceroceras
brongniarti (J. Sowerby), Bajocian, Spherocone, aperture depressed,
without rostrum or lappets. (After Quenstedt. ) /, /', Baculites (Cyrto-
chilus) baculoides (Mantell), Cenomanian, Chardstock (Somerset).
(After Crick. ) All x |, except where otherwise stated. K, rostrum ;
L, lappet ; S, constriction,

160 PALEONTOLOGY

project outwards as a horn, or even be curled backwards
(Fig. 52). Genera with a prominent rostrum are found
onwards to the Cretaceous period, but in certain Lower
Jurassic forms a pair of lateral protuberances (lappets)
appear, and as those increase the rostrum diminishes
until it may disappear altogether (Fig. 46, a-h). Genera
with large lappets are confined to the Upper Jurassic.
In many cases the lappets tend to close in the aperture,
the extreme case being that of Morphoceras (Fig. 46, f),
where the aperture is even more completely subdivided
than in the Silurian Tetrameroceras.

The constrictions already described in Gastnoceras are
found in various other ammonoids, sometimes at intervals
throughout life (Fig. 41, d), sometimes only in very early
stages of growth. As we have seen in gastropods, all
specializations of the apertural margin must have inter-
fered with the simple growth of the shell, and we must
suppose, either that they were not formed until the shell
had attained its full size, or that, if formed, they had to be
resorbed before growth could continue. Either of these
suppositions may be true in particular cases. ' In a
number of genera constrictions do occur at intervals
throughout the length, and in one case (Perisphinctes and
allied genera) there are sometimes found peculiar
markings (parabola) at intervals which may indicate the
places where lappets were resorbed.

In the body-chamber of some ammonites there is
frequently found a pair of symmetrical calcareous plates,
which together form a heart-shaped body (aptychus).
Similar aptychi are found apart from ammonites, though

THE CEPHALOPODA 161

in strata in which they occur, and there are even
certain Jurassic shales in the Alps which are full of
aptychi, but contain no ammonites. As the former are
made of calcite, while the shells are aragonite, more
easily destroyed by solution, this is not surprising. In
one unique example, from the Inferior Oolite of Dundry
Hill, Somerset, the aptychus was found in its natural
position, closing in the aperture of the shell, so that it
evidently acted as an operculum. The two halves of
the aptychus were permanently united in Scaphites
(synaptyclws) ; in the A vietidce and A maltheidce and some
Palaeozoic genera, a single horny plate (anaptychus)
served as an operculum, and many ammonoids possibly
had no operculum of any kind.

The septa of ammonoids always show a tendency to
folding at the margin, giving rise to more or less com-
plexity in the suture-line. We know nothing of the
meaning of this folding, puckering, and frilling, but
the various resulting patterns of suture-line are of
the greatest value in tracing affinities and classifying
ammonoids, so that careful attention must be paid to its
details. The number of lobes and saddles is extremely
small at first (Fig. 47), but increases during the newer
Palaeozoic era, some families then having a very large
number (Fig. 47, e, /). In the typical families of the
later Mesozoic the number tends to keep within certain
limits, and to bear a close relation to the degree of in-
volution of the shell. The following description applies
especially to those genera in which this relation holds,
and needs some modification to apply to the Palaeozoic

ii

1 62

PALEONTOLOGY

FIG. 47. — SEPTAL SUTURES OF EARLY AMMONOIDEA.

Each figure has the umbilical margin on the left and the middle line of the
periphery on the right. The arrow points towards the body-chamber.
a, Anarcestes, Devonian ; b, Manticoceras, Devonian ; c, Tornoceras,
Devonian ; d, Gastrioceras, Carboniferous ; e, Prolecanites , Carboni-
ferous;/, Medlicotta, Permian; g, Paraceratites, Triassic; h, Phyl-
loceras, Jurassic, a-/, After Crick ; g, after Perrin Smith ; h, after
d'Qrbigny.

THE CEPHALOPODA 163

forms with either very few or very many lobes and
saddles.

Beginning at the periphery, we find there a median
external lobe (also called peripheral or siphonal), usually
divided in the middle by a small saddle. Flanking this
lobe on either side, at about the peripheral margins, is a
pair of external saddles. On each lateral area are two
lobes and two saddles (first or superior, and second or
inferior lateral). If we are dealing with an evolute shell
there remains only a median internal lobe (also called
dorsal or antisiphonal). In the case of shells tending
towards the involute, additional lobes and saddles are
developed below the lateral lobes : these are called
auxiliary lobes and saddles, and their number increases
with the degree of involution, though not at the same
rate in different families.

The degree of complexity of the suture-line serves as a
general indication of age. Simple, undivided lobes and
saddles indicate the Devonian or Carboniferous period
(Fig. 47, a-e] ; in the Permian, subdivision begins
(Fig. 47, /). The Triassic period is specially character-
ized by ammonoids with broad rounded saddles and
denticulate lobes (ceratitic sutures, Fig. 47, g), but along-
side these are others with highly complex sutures. In
the Jurassic period the ceratitic type is absent, and the
highly complex (ammonitic) suture is alone found (Fig. 47,
h, 48, a-h). This latter also characterizes the Creta-
ceous period, but in sub-tropical and tropical regions
there are Cretaceous forms (pseudo-ceratites) in which
the suture-line reverts by catagenesis to something like
the common Triassic type (Fig. 48, *,/).

164

PALAEONTOLOGY

FIG. 48. — (For description see p. 165.)

THE CEPHALOPODA 165

The shape of the whorl-section varies greatly : it may
be roundfd, quadrate, cordate, sagittate, depressed, etc. (Fig. 49).
These various forms, in combination with different degrees
of involution, of rapidity of growth, and of change with
advancing age, give rise to an enormous variety of shell-
forms, to some of the commonest of which special names
are given. Thus, when increase of size is slow, so that
for a given diameter there are many whorls, and there is
a wide, shallow umbilicus and a square periphery, the
resemblance to a cart-wheel suggests the term rotiform ;
but with a rounded periphery, as though the wheel were
pneumatically-tyred, the term planulate is used (Fig. 49,
a, b). Discoidal means much compressed with an acute
periphery, usually with very small umbilicus. When the
whorls are depressed, with broad periphery and rather
deep umbilicus, the ammonite if placed on its side suggests
the shape of a crown, especially if (as is common) it has
a row of tubercles or spines along the lateral margin :
such an one is coronate (Fig. 49, /). Mr. Buckman uses
the term oxycone for discoidal forms and those approaching
them (Fig. 49, c, d), this being a stage which corresponds
to the sphaerocone but differs from it in being com-
pressed ; and cadicone for the extreme form of the coronate
type, with very deep conical umbilicus (Fig. 49, i).

FIG. 48. — SEPTAL SUTURES OF JURASSIC AND CRETACEOUS
AMMONOIDEA.

Arrangement as in Fig. 47. Circles in e denote the position of tubercles.
a, Paltopleuroceras, Domerian ; b, Grammoceras toarcense, Yeovilian ;
c, Oppelia, Upper Jurassic; d, Caloceras, Hettangian; e, Deroceras,
Charmouthian ; ft Cosmoceras, Callovian ; g, Hoplites, Lower Cre-
taceous; h, Perisphinctes, Upper Jurassic; i, Tissotia, Upper Cre-
taceous; /, Metengonoceras, Upper Cretaceous, a-h, after d'Orbigny;
i, after H. Douvill£ ; /, after Hyatt.

1 66

PALEONTOLOGY

FIG. 49.—WHORL-SHAPE IN AMMONOIDEA.

All except e, g, and h are apertural views, a, Microderoccras, lat-umbili-
cate, rotifortn, whorl-section quadrate, b, Dactylioceras. lat-umbili-
cate, planulate, whorl-section rounded, c, Chamousettia, angust-

THE CEPHALOPODA 167

The ornamentation of ammonoids falls under a few
definite heads. Primitive genera are smooth, and so are
many higher forms in early youth. The smooth stage is
generally followed by a capillate stage, in which there is a
pattern of fine raised lines on the surface ; and this by
subcostate and costate stages, in which the much more
prominent features called ribs (pilce, costce) are developed :
these are really corrugations of the shell, and usually
appear as strongly on the cast as on the exterior ; when
they do not, it is because an internal partition or septum
is formed under each rib, as in the planulate ammonites
of the Upper Lias. A still later stage is the tubenulate
or spinous, when portions of the ribs stand out promi-
nently either with rounded or pointed ends. These
stages follow one another in the given order (anagenesis
of ornament), though any one of the intermediate stages
may be skipped in the hurrying-on of development.
After the last stage, catagenesis may often take place,
and the shell which is tuberculate in youth may become
costate or even smooth in old age.

The ribs, or capillae, run in a straight or curved direc-
tion from the umbilicus to the periphery ; their course is

FIG. 49. — WHORL-SHAPE IN AMMONOIDEA (continued).

umbilicate, lenticular oxycone, whorl-section cordate, d, Amaliheus,
angust-umbilicate, discoidal oxycone, whorl-section sagittate, e, Dia-
gram based on Aspidoceras to explain areas of whorl-surface. /, Erym-
noceras, coronate, whorl-section greatly depressed (periphery and
umbilical area practically meet, excluding any lateral area), g, Peri-
sphinctes, section showing early whorls of depressed crescentic shape,
changing later to compressed and planulate. h, Quenstedtoceras,
section showing early compressed whorls, changing later to broad and
somewhat depressed, i, Cadoceras, a cadicone, greatly depressed
whorls, deep conical umbilicus (indicated by dotted lines), a, b, c, d,
f, i, After d'Orbigny; g, after Dumortier and Fontannes ; h, after
R. Douville".

1 68 PAL/EONTOLOGY

constant through the various stages of development, and
is a good guide to the affinity between -different species.
This course is termed the radial line of the species, and it
bears a close relation to the form of the apertural border.
Thus, if there is a rostrum and lappets without any con-
striction behind, the radial line will be more or less
sickle-shaped or falciform (Figs. 46, /, h\ 47, d), but if
there is a constriction it will be straight (Fig. 47,7)- A
rostrum without lappets gives rise to a radial line like an
inverted L, or Greek T (gammi-radiate, Figs. 39, a\ 45, d, e).
The ribs sometimes cross the periphery quite straight
(Fig. 46, h, k) ; more often with more or less of a forward
bend, owing to the general presence of a rostrum (Fig.
45, /). The absence of a forward bend in ttfe ribs does
not, however, prove the absence of a rostrum (Fig. 46, a) ;
it may mean that rostra were produced at intervals and
resorbed. In many forms the ribs are interrupted on the
middle of the periphery, either by an elevated keel (carinate
periphery, Fig. 39., b) or by a median groove (sulcate
periphery, Figs. 45, d\ 46, *') or by a keel flanked by
grooves (carinati-sulcate] .

The radial ornament is in some families crossed by
longitudinal lines.

The classification of the ammonoids is by no means in
a settled state yet, and it is necessary to have some
acquaintance with the development of ideas on the sub-
ject. A hundred years and more ago, all ammonoids
whose shells were normally coiled with the whorls in
contact were given the one generic name Ammonites,
while those which showed peculiar modes of coiling

THE CEPHALOPODA 169

received special names such as Hamites, Scaphites, Turri-
lites, and Baculites. Leopold von Buch in 1830 pointed
out the importance of the suture-line and separated from
Ammonites the Palaeozoic species as Goniatites and the
most typical Triassic species as Ceratites. D'Orbigny
added several to the list of genera based upon peculiari-
ties of coiling. This classification remained unchanged
until 1865 : it was inconsistent and lop-sided. Based
mainly on coiling — a feature that may vary during the
life of the individual — it yet united under the one name,
Ammonites, species with all degrees of involution,
from those in which the whorls only just touched to
those which were completely involute, yet it separated
off under a distinct name those in which the whorls just
failed to touch. The number of species of Ammonites
too was too enormous, whatever conception might be
held of a genus.*

Von Buch had indeed divided Ammonites into some
sixteen groups, but these were based upon external form
and ornament only ; while Montfort (1808) and de Haan
(1825) had separated off certain groups under distinct
generic names, but these had been ignored. In 1865
Suess in Europe, and in 1867 Hyatt in America, inde-
pendently and almost simultaneously started the sub-
division of the genus Ammonites. Their method of work
was rather different. Suess (whose work was quickly
carried on by Waagen and Neumayr) defined a few large
natural groups and gave them generic names. Hyatt,

* Nevertheless, if used colloquially, the word "ammonite," in
the sense of the old generic name, will always remain a useful one

170 PAL/EONTOLOGY

inspired and started on his work by Louis Agassiz, con-
fined himself to the ammonites of one small division o
time and divided them with much greater minuteness.
Thus it happened that one of Waagen's genera (Arietites)
was equivalent to a whole family of Hyatt's. The
American worker also initiated the method of using
individual development (ontogeny) as an essential criterion
in classification. This double start led to some confusion
in the naming of Jurassic ammonites, which Mojsisovics
and Gemmellaro, working on the almost untouched mass
of Triassic and Permian species, were able to avoid.
Numerous later workers have added greatly to the recog-
nized number of Ammonite genera — work which is not
always greeted by stratigraphers with the thanks which
it deserves, but which is inevitable once the evolutionary
conception of a genus is accepted.

GENERAL HISTORY OF THE AMMONOIDEA.

As so often happens, the apparently most primitive
genera are not the earliest. The orthocone Bactrites
indeed is doubtfully reported from the Silurian, but is
not known in the Devonian until nearly the end of the
Middle epoch of that period ; while Gyroceras [Mimoceras],
which starts as a gyrocone and continues as an ophiocone,
is known from the Middle Devonian only. At the very
beginning of the Devonian there appear two genera in the
ophiocone stage, Anarcestes (Fig. 47, a) and Agoniatites
(Fig. 41, b), the very simple sutures of which have led
most palaeontologists to place them in a family Nautilinida,
though they might perhaps better be regarded as at the
base of separate other families. Towards the end of
Middle Devonian time, a new family Magnosellaridce

THE CEPHALOPODA 171

appears (Tornoccras, etc.), characterized by the appearance
of small external and large lateral saddles (Fig. 47, c). In
the Upper Devonian flourished the remarkable family
Clyweniida, in which the siphuncle is on the inner or
dorsalmargin, instead of the peripheral margin. In the
same period lived the Gephyvoceratidce and Magnoseliarida
(typified respectively by Manticoceras and Tornocems, al-
ready described). With the beginning of the Carboni-
ferous period the Devonian genera vanish, while two
new families appear — the Glyphioceratidce in which the
peripheral lobe comes to be divided by a small saddle,
and surface-ornament first begins to be noticeable
(GlyphioceraS) Gastrioceras) and the Prolecanitida in which
the number of lobes and saddles is much increased, and
they come to have a narrow tongue-like shape (Pfolecanites,
Fig. 47, e) . This family attains much greater complexity of
suture in the Permian, especially in Medlicottia (Fig. 47, /)
and in Cyclolobus and Popanoceras, which seem to be tran-
sitional from this family to .the Triassic Arcestida. The
type-genus of this last, Avcestes (Fig. 45, c), is notable for
the extreme length of body-chamber, which, in combina-
tion with the U-shaped whorl-section due to combined
depression and involution, must have resulted in a most
extraordinarily shaped animal.

Along with these in the Trias are found the families
Cevatitida and Meekoceratida (the latter also Permian,
Fig. 47, g), in which the suture-line is generally " cera-
titic," or in some genera rather more complex. The
former family attains to costate and tuberculate ornament,
but the latter are smooth, compressed, discoidal forms.
In the Trachyceratidce (Fig. 45, d) the surface becomes
ornamented by intersecting radial and longitudinal lines,
with tubercles at the intersections, and the periphery is
more or less sulcate. The suture-line is ceratitic, but
with saddles notched somewhat as mAsteroceras. Another

172 PAL/EONTOLOGY

family in which involute, compressed forms with acute
periphery predominate is the Pinacocevatidce, in which the
suture ranges from ceratitic to the most highly complex
kind known in any arnmonoid. These are the principal
Triassic families, none of which survive that period.

The classification of the Jurassic ammonites is in a
very unsettled state at present. Mr. Buckman, the chief
English investigator of them, proposes the following
provisional scheme (the terms involved are explained
farther on) :

I. Suture-line phylloidal, no keel, no aptychus or anap-

tychus. Families Phylloceratidtf, Lytocevatidce.

II. Suture-line phylloidal, passing to complex and much

inflected ; no keel in most, when present it arises
late, after elaboration of suture-line ; with aptychus
or anaptychus.

III. Suture-line with saddle of ceratitic outline, more or
less serrated but generally remaining fairly simple ;
keel usually developed at a very early stage ; elabora-
tion of suture-line, if any, coming later ; with aptychus
or anaptychus.

I. While most of the Mesozoic families of ammonites
are short-lived, rapidly attaining their acme and dying
away in a single geological age or little more, the two
families of this branch are exceptional in enduring with
little modification through the Jurassic and Cretaceous
periods. They are also peculiar in having a narrower
geographical distribution than most other families,
their headquarters being in the subtropical waters of
" Tethys " — the ocean which occupied the site of the
present Alpine- Himalayan mountain-systems and the
Mediterranean region. Only at intervals of time did a
few representatives succeed in migrating into the colder
seas of what is now Central and Northern Europe. The

THE CEPHALOPODA 173

Phylloceratidae are distinguished by their suture-line
(Fig. 47, h), in which the saddles are subdivided by little
lobes (lobules) into ovoid " cells," constricted posteriorly
(phylloids) ; by the general absence of ornament other
than striae ; and by the general tendency to complete
involution in the shell (Fig. 46, a).

The Lytoceratidae also have a " phylloidal " suture,
but it often becomes so complex that this character
is obscured ; long lateral branches to the median internal
lobe are characteristic. They tend to an evolute form,
though with exceptions, and in the Cretaceous period
yield a large proportion of the uncoiled forms (Macvo-
scaphites, Ancyloceras, some species of Criocevas}. In
both these families constrictions may occur at intervals,
and in the Lytoceratidae these are sometimes marked on
the exterior of the shell not by depressions, but by con-
spicuous elevations (" flares ").

II. and III. It will be convenient to describe the rest
of the Jurassic ammonites in historical order, indicating
their taxonomic position by the number II. or III. in
parentheses.

The Jurassic period opens with the Hettangian age,
characterized by the first appearance of ammonites in
Northern Europe. These belong to the family Psilo-
ceratidae (II., Fig. 48, d). The genera vary from nearly
smooth (Psilocevas) to strongly costate (Schlotheimia). The
latter genus survives to the end of the Sinemurian age,
which is mainly characterized by the Arietidae (III.), a
family with uninflected sutures, simple straight ribs bend-
ing forward on the periphery (Asteroceras, Fig. 39, etc.)
and a keel. The whorl-section is usually quadrate, but in
the degenerate Oxynoticevas it becomes acutely sagittate :
this and similar oxycones persist into the next age. Along-
side of the Arietidae in their acme are found not only some
survivors (Schlotheimia} from a previous age, but also the

174 PAL/EONTOLOGY

primitive ancestor of forms which attain their acme
later — Microderoceras Urchi being the first. This is a
widely umbilicate planulate form, whose inner whorls
show anagenesis from smooth to costate and tuberculate.
Xipheroceras (p. 134) is an allied form, later in time.
From some such form may have sprung the Derocera-
tidae (II.), which attain their acme in the Charmouthian
stage — forms with rounded whorl-section, costate tuber-
culate or spiny, and a highly ornate, much inflected suture
(Fig. 48, e). Alongside these are found two other families
— the Liparoceratidae (III.), which begin as "capricorns "
much like Xipheroceras, and rapidly develop into multi-
tuberculate unkeeled sphaerocones with uninflected
suture-line, and the Polymorphidae (II., Fig. 46, e), with
an inflected suture, which also begin rather like capri-
corns, but become bi-tuberculate (two tubercles on each
side) and develop a keel later on. Here also appear
Cceloceras pettos, the first Jurassic coronate type, ancestral
to the dominant families of the Upper Jurassic ; and
•Eckioceras, a genus which like the Polymorphidae passes
from a planicostan to a keeled stage, and may be ancestral
to the Hildoceratidae so important a little later.

In the Domerian stage, there is an irruption of a new
family, Amaltheidae (III.), which have several features
in common with the Arietidae, but in which discoidal
forms predominate, and the keel has typically a knotted
appearance (Figs. 45, e ; 48, a ; and Frontispiece).

The Whitbian, Yeovilian and Aalenian stages are
dominated by the enormously prolific Hildoceratidse (III.),
which are already abundant in the Domerian of the
Mediterranean region, though rare in Britain. They are
characterized by a more or less doubly-curved or sickle-
shaped radial line of ornament ; the whorl-shape and
degree of involution vary greatly, but there is always a
keel, and the suture-line is uninflected, but differs from

THE CEPHALOPODA 175

that of Arietidae in the much deeper first lateral lobe
(Figs. 45, /, h\ 46, ft, c; 48, b).

In the Whitbian, the hildoceratids (Harpoceras, Hildo-
ceras, Pseudolioceras) are accompanied by planulate dero-
ceratids, some tuberculate (Peronoceras, Porpoceras), others
degenerated to costate — the dactyloids (Dactylioceras). In
the Yeovilian and Aalenian, other hildoceratids (Lioceras,
Ludwigia)axQ accompanied by the last polymorphids, some
closely resembling hildoceratids in aspect (Dumortieria),
others quite distinct (Tmetoceras).

In the Yeovilian (sporadically), the Aalenian (com-
monly), ending in the earliest Bajocian, occur the Ham-
matoceratidae (II.), tuberculate with rounded periphery,
and deroceratid suture-line, which develop a keel and
acquire striking resemblance to hildoceratids and
sonninines, from which they are distinguished by the
suture-line.

As we pass from the Aalenian to the Bajocian comes
the greatest faunal change of the Jurassic : here the line
between Lower and Upper Jurassic epochs should be
drawn, though the old stratigraphical boundary between
Lias and Oolites comes at least an age earlier. The
hitherto dominant families die out suddenly, and three
new series rapidly establish themselves. First are the
Sonnininse (III., Fig. 46, /), a sub-family of the
Amaltheidae, though separated from it, so far as present
knowledge goes, by nearly three ages. They are
characterized in general by a catagenesis of ornament,
the inner whorls being spinose, the outer plainly
ribbed or even smooth (a marked contrast to the
anagenetic ornament of the Domerian amaltheids), but
the suture-line becomes more elaborate as the ornament
declines. The keel is in many cases a septicarina
(p. 147). The Sonnininae are almost confined to the
Bajocian ; but the other two new families last to the end
of the Jurassic period.

1 76

PAL/EONTOLOGY

One of these is the Oppelidye (HI.), keeled forms
which appear suddenly in the oxycone stage, without
sign of preceding stages, and mostly remain in that stage
or tend to open out into scaphitoids ; they tend to produce
well-marked lappets and a rostrum, so that their radial
line is sickle-shaped (Figs. 46, d\ 50) ; their suture-line has
a smooth outline and a very prominent first-lateral saddle
and deep first-lateral lobe (Fig. 48, c).

The other is the family Stepheoceratidae (II.), using

FlG. 50. — LONULOCERAS BRIGHTI (PRATT), CALLOVIAN (LOWER

OXFORD CLAY), CHRISTIAN MALFORD (WILTS).
(Natural size.)

Rostrum and lappets. Original.

that term in its widest sense : a very extensive group,
which starts from a coronate stage, and goes through
various developments of form — some becoming planulates
directly, others sphaerocones (Fig. 46, k), others planulates
through a sphaerocone stage, but all retaining the same
general type of ribbing — more or less straight ribs,
splitting on the lateral area into two or more which
either run contiguously across the arched periphery
(Fig. 46, h) or are interrupted by a groove (Fig. 46, *'),
never a keel. The aperture may have a plain border,

THE CEPHALOPODA 177

or there may be lappets (sometimes long and elaborate
in shape) but they never affect the course of the radial
lines, though sometimes they leave indications of their
former position by markings (parabolic knots) at intervals.
Constrictions occur in some genera (Fig. 46, /, /£). These
richly varied Bajocian and Vesulian stepheoceratids are
continued into higher Jurassic stages chiefly as costate
planulates known as the Perisphinctidae (Figs. 46, j;
48, h ; 49, g) — certainly not a true family, as they are
polyphyletic, arising from several different coronate
stocks.

The Bathonian age in Northern Europe was marked
by an almost total disappearance of ammonites, but in
the Callovian a rich fauna reappears, in which four new
families appear alongside the persistent oppelids and
perisphinctids. These are (i) the Cardioceratidae, pre-
sumably allied to Oppelidae : they pass from platycones
to sphaerocones with varying degrees of inflation (Quen-
stedtocems, Fig. 49, h, Cardioceras), even to very stout
cadicones (Cadoceras, Fig. 49, *) ; (2) the Pachyceratidae,
coronates ; (3) the Aspidoceratidae, at first square- whorled
and evolute, later (in some genera), rounded and involute,
characterized by a modification of the coronate or peri-
sphinctid suture, in which the external saddle and first
lateral lobe become very large, the lower saddles and lobes
being compressed into what looks like a single complex
saddle ; (4) the Cosmoceratidae, starting as closely-ribbed
coronates (Kepplentes) but rapidly becoming more com-
pressed and developing a periphery medianly smooth,
often with marginal tubercles, while the lateral area
becomes richly bi-tuberculate : long lappets are present
(Figs. 46, g; 48,/). This last family hardly outlasts the
Callovian, and the Cardioceratidae die out in the Kim-
meridgian ; the remainder survive to the end of the

Jurassic,

12

178

PALEONTOLOGY

The Phylloceratidae and Lytoceratidae persisted in the
Alpine-Mediterranean region right through the Cretaceous
period. The latter family gives rise to some uncoiled
forms (Macroscaphites, Ancyloceras, Fig. 41, h, i) in the
Lower Cretaceous, but also giving off a branch family,
Desmoceratida;, in which the tendency is towards tighter
coiling or involution : this family lasts to the Upper
Cretaceous.

In the Neocomian stage of the Lower Cretaceous, the
oppelids and perisphinctids persist, and a very charac-
teristic form is Olcostephanns (also found in the higher

FIG. 51. — OLCOSTEPHANUS ASTIERIANUS (D'ORBIGNY)
NEOCOMIAN. ( x ^.)

Plain apertural margin, oblique to radial line of ornament. (After
d'Orbigny. )

Jurassic strata), coronate to planulate, with perisphinctid
aspect (Fig. 51), but a different suture-line of the
" inverse " type (the lobes from periphery to umbilicus
successively ending farther and farther in front of the
guide-line). Besides the uncoiled forms already mentioned,
Cnoceras and Toxocems (Fig. 41, /) are common, or more
correctly criocones and toxocones, for the generic names
as generally used denote groups of homo2omorphs rather
than true genera.

The Barremian stage is specially characterized by the
Pulchellida, rather compressed involute forms with ribs
of peculiar shape — broadening from umbilicus to peri-

THE CEPHALOPODA 179

phery, so that the spaces between them are parallel-sided
and at last narrower than the ribs themselves : the ribs
form sharp tubercles on the edge of the flattened periphery.

In the Aptian and Albian stages the family Hoplitidce
is abundant : it is regarded as derived from the peri-
sphinctids, though it differs greatly from them both in
aspect and suture-line (Fig. 48, g). These ammonites
are rather involute, with flattened periphery, and sigmoid
(doubly curved) ribs. The later species tend to be strongly
bituberculate, with the ribs gathered in to the tubercles.
Such species abound in the Lower Gault Clay of
Folkestone, but in the Upper Gault they have almost com-
pletely disappeared and are replaced by keeled forms
practically unknown in the Lower Cretaceous. The
uncoiled forms most characteristic of the Gault or
Albian stage are Hamites and Turrilites (Fig. 41, e,j).

An important family beginning in the Lower
Cretaceous but extending into the Upper is the Acantho-
ceratidw, characterized by broad straight ribs with a
tendency to break up into many tubercles. Characteristic
of the Upper Cretaceous are the keeled forms of the
family Pvionotvopidce (Mortoniceras, Fig. 52 ; SMcenbachia).
In warmer latitudes there appear a number of genera
known as pseudo-ceratites (Fig. 45, i, j), not a natural
family, but agreeing in a degeneration of the suture-
line which reverts to something resembling the ceratitic
type of the Trias. These forms are known in Southern
Europe, Syria, India, North and South Africa, South
America, and the Southern United States. The cata-
genesis of suture is not accompanied by any loss of
coiling: many of them are strongly involute. On the
other hand, the uncoiled Cretaceous genera show complex
sutures. Scaphites (Fig. 41, g) and Baculites (Fig. 46, e)
are most characteristic of the Cretaceous beds higher than
the Gault.

i8o PALAEONTOLOGY

There are certain important lessons which the study
of the enormous number of species of ammonites has
impressed upon palaeontologists. First is the danger of
basing any idea of relationship on external form alone.
Over and over again species of very similar form have
proved to differ widely either in their suture-details, or

FlG. 52. — MORTONICERAS INFLATDM (J. DE C. SOWERBY),

ALBIAN (UPPER GAULT.) (xf.)

Ornament omitted in centre; bifurcating ribs replaced later by single
multituberculate ribs. Rostrum bent backwards. (After Buvignier. )

in their ontogeny (as shown by a study of the inner
whorls) : such species are said to be homceomorphs of one
another, and have more than once led, either to mistakes
in zonal stratigraphy, or to the erroneous idea that
certain ammonite-species have too long a range in time
to be of zonal value.

Secondly, in spite of the richness of the ammonite-

THE CEPHALOPODA 181

faunas, the gaps in the phylogenetic sequence are im-
pressive. Neumayr, forty years ago, first called special
attention to this in a famous essay. He pointed out how
at certain geological horizons there appeared abundant
new ammonites which could not be the descendants of
any in the earlier zones : these he termed crypto genetic
types. In other cases, similar forms reappeared at wide
intervals, while absent from intervening strata : these he
termed sporadic types. Two particular cases of the
latter he was able to explain very satisfactorily : at
various horizons in the Jurassic of South Germany (and
at some but not all of the same horizons in England)
there appear species of the genera Phylloceras and Lyto-
cevas (or, as we should now express it, of the families
Phylloceratidae and Lytoceratidae). But similar forms
occur throughout the Jurassic of the Mediterranean
region ; hence Neumayr's explanation of their sporadic
occurrence farther north was that opportunities for migra-
tion occurred occasionally owing to geographical or
climatic changes, and he expressed the belief that other
sporadic and cryptogenetic types would prove to have
migrated from unknown regions. Since Neumayr wrote
it has been shown that many of his supposed " sporadic "
types are not related to one another but are homoeo-
morphs, as for instance various oxycones which at that
time were placed on account of their form in the genus
A maltheus.

Cryptogenetic types are indications of the imperfection
of the palaeontological record, which, as Mr. Buckman
points out, may be due to several causes :

182 PALEONTOLOGY

1. Incomplete exploration of large areas of the earth's
surface — of which indeed three-quarters is practically
preserved from exploration by a covering of deep sea.

Within the last twenty years, exploration of new areas
has actually revealed the ancestors of what had been
cryptogenetic types (not indeed of ammonites, but of
trilobites and mammals), and justified Neumayr's migra-
tion-theory.

2. Destruction of fossiliferous strata by denudation.
Not only have large parts of some continents been swept
bare of masses of strata that once covered them (as
proved by isolated patches preserved by some accident),
but it is now realized that among shallow-water deposits
there has been much " pene-contemporaneous " erosion,
and that in what looks like a thick conformable series of
sediments there may be many gaps due to this cause.

3. New stocks begin with small members which are
liable to be overlooked in collecting or in subsequent
investigation ; and even larger forms, if rare (and they
may be rare if the stock has not yet attained dominance),
are liable to be neglected.

DIBRANCHIATA.

The most familiar fossils of this order are those known
as Belemnites, the most typical of which are found in the
Jurassic and Lower Cretaceous systems, though forms
but little different also occur in the Upper Triassic and
Upper Cretaceous. A typical belemnite shell consists of
(i) a phvagmocone (chambered shell) which is an orthocone
or cyrtocone with a calcareous globular protoconch,

THE CEPHALOPODA 183

a siphuncle along the ventral margin and a forward
prolongation of the dorsal region, the pro-ostracum
(Fig. 53, a, b, b') and (2) a solid guard, more or less
cigar-shaped, with a conical hollow (alveolus] at the
blunt end for the reception of the apex of the phragmo-
cone. The guard is composed of fibrous calcite arranged
radially ; it is more often found preserved than the
phragmocone, and is so resistant to wear and tear that
belemnite guards are often found as derived fossils in
gravels. In a few cases, sufficient traces of the body of
the belemnite have been found to show that the skeleton
was entirely internal, the animal growing out of all pro-
portion to the growth of the shell, so that an ever smaller
portion of the body remained within the body-chamber,
and the mantle gradually enveloped the whole shell.
Like its relative the modern cuttle-fish, it possessed an
ink-sac and a number of arms around the mouth, though
these were furnished with horny hooks not known in the
cuttle-fish, and there seem to have been only six of them
instead of the eight or ten of modern dibranchiates
(Fig. 53, a).

In some of the earliest (Upper Triassic and Lower
Jurassic forms) the phragmocone extends nearly to the
end of the guard (Fig. 53, ' c\ and in these cases it is
probable that the former was still an external shell in
early life and only later became enveloped in the mantle,
which then began .to secrete a guard around it. (This
makes the name "guard" inappropriate, since it was
formed when the phragmocone least needed protection).
But in most later forms the alveolus only extends a

i84 PALEONTOLOGY

short distance into the guard, which probably began to
be formed as early as the phragmocone.

The phragmocone when found alone is apt to be
mistaken for an Orthoceras, but it can be distinguished,
(i) by the marginal position of the siphuncle, and (2) by
the fact that the growth-lines (conothecal since] , which
form its only ornament, curve forwards on the dorsal
region in correlation with the presence of a pro-ostracum

(Fig. 53. 4

The chief features serving to distinguish speciesiamong
belemnite-guards are (i) 'the general shape, (2) the dis-
position of grooves on the surface, (3) the depth and
apical angle of the alveolus. The two main shapes are
the lanceolate and the hastate : in the former the diameter
is constant for the greater part of the length, and the
posterior end is conical; in the latter the diameter in-
creases from the front end for some distance backwards.
In cross-section either form may be circular, compressed
(laterally) or depressed (dorsi-ventrally) : to distinguish
between the latter in the absence of the phragmocone, it
is sufficient to notice the plane of symmetry of the
alveolus which is the median plane. The alveolar angle
varies between 12° and 32°.

A few belemnite-guards are quite free from grooves ;
others have a number of short apical grooves ; but the
majority have either a median ventral groove or a pair of
dorso-lateral grooves.

The classification of belemnites is in a confused and
unsatisfactory condition. Only morphological classifi-
cations have been proposed, except for limited groups of

THE CEPHALOPODA

185

arms

FIG. 53. — BELEMNITES.

a, Restoration of the living Belemnite (modified from d'Orbignyf).
F, Funnel ; G, guard ; Pro., pro-ostracum. b, b' , Restoration o
phragmocone, in front view and section (after Crick). P, Protoconch ;
Phr., phragmocone; S, siphuncle. ct Atractites philippii Hjatt and
Smith, Carnic, California. Guard (G) partly broken open showing
phragmocone (Phr.}. (After Hyatt and Smith.) d, B.ekmnofsis
bessinns (d'Orbigny), Bathonian. e, Hibolites hastaius (Blainville),
Callovian (Oxford Clay). /, Hastites clavatus (Blainville), Char-
mouthian. g, Oxyteuthis acuius (Miller), Sinemurian. h, Coelo-
teuthis excavaius (Phillips), Sinemurian. d-h, Ventral views : dotted
lines indicate the alveolus, unbroken lines the grooves on the surface.
d'-h', Cross-sections, the dot marking the axis, d, After d'Orbigny ;
e-h, after Phillips. (All X |, except a).

186 PALEONTOLOGY

species, and little notice has been taken of ontogeny,
although the mode of growth of the guard (by accretion)
makes changes in shape and grooving recognizable.

The following is an imperfect statement of the classi-
fication most generally accepted. An asterisk denotes
genera always accepted as distinct, the remainder being
usually lumped in the comprehensive genus Belemnites.

A. GUARD NOT GROOVED.

Atrocities* Trias, Scythic to Rhaetic (guard nearly
cylindrical, alveolus nearly as long as guard, Fig. 53, c).

Oxytenthis, Sinemurian-Charmouthian (guard acutely
conical, alveolus half length of guard), e.g. Bel. acutus
(Fig- 53> g)-

B. GUARD WITH VENTRAL GROOVE.

C&loteutkis, Sinemurian (short, cylindro-conical, alve-
olus nearly as long as guard, groove very wide), e.g. Bel.
excavatus (Fig. 53, h).

Belemnopsis, Charmouthian-Bathonian (long, cylin-
drical, with acute apex, groove long, deep and narrow,
alveolus one-third length of guard), e.g. Bel. bessinus
(Fig. 53, d).

Pachyteuthis, Charmouthian-Neocomian (stout, cylin-
drical, with conical apex, groove long and shallow,
alveolus from one-third to two-thirds length of guard),
e.g. Bel. abbreviates (Fig. 54, a).

Cylindroteuthis, Callovian-Neocomian (long, cylindrical,
groove deepest near apex, becoming indistinct forwards,
alveolus about one-fourth length of guard), e.g. Bel. oweni
(Fig. 54, ft).

Hibolites, Aalenian-Cenomanian (hastate), e.g. Bel.
hastatus (Fig. 53, e).

Belemnitella,* Emscherian-Maestrichtian (cylindrical,
with rounded and acuminate apex, alveolus one-third
to one-half length of guard, ventral slit in guard along

THE CEPHALOPODA 187

more than half the length of alveolus), e.g. Bel. mucronata
(Fig. 54, 4

Actinocama%* Cenomanian-Campanian (very like the
last, but the part of the guard around the alveolus is im-
perfectly calcified, and is always more or less destroyed),
e.g. A. quadratus (Fig. 54, d).

C. WITH DORSO-LATERAL GROOVES.

Aulacoceras* Trias, Ladinic to Noric (phragmocone
much longer than the small hastate guard).

Hastites, Charmouthian-Neocomian (hastate), e.g. Bel.
clavatus (Fig. 53,/).

Pseudobelus, Aalenian-Cenomanian (lanceolate, with
very deep lateral grooves), e.g. Bel. Upartitus.

D. WITH SEVERAL APICAL GROOVES.

Dactyloteuthis, Domerian-Yeovilian (cylindro-conical
up to a certain age, then suddenly elongating into a long
tubular apex), e.g. Bel. acuarius (Fig. 54, /).

Megateitthis, Aalenian-Bajocian (with similar change
of shape), e.g. Bel. giganteus.

E. WITH MEDIAN DORSAL GROOVE.

Duvalia, Callovian-Neocomian of Mediterranean pro-
vince (much compressed), e.g. Bel. dilatatus (Fig. 54, e).

The typical Belemnoids die out at the end of Meso-
zoic time, except in Australia, where they are recorded
from Eocene strata. In Europe, there are in the Eocene
several forms, mostly very rare as fossils, which may be
their descendants strangely modified before final extinc-
tion. Greater interest attaches to two collateral stocks
which lead to living forms. One of these is the Eocene
Belosepia, which forms a link between the belemnites and
the recent cuttle-fish Sepia. The siphuncle is greatly
widened, the phragmocone is transitional from its typical
form to the "cuttle-bone," and the guard is in process

i88

PALEONTOLOGY

of reduction to the little point of a modern Sepia. The
other transitional form is the Miocene Spirulirostra, in

FIG. 54. — BELEMNITES.

a, Pachyteuthis explanaloides (Pavlow), Upper Jurassic, b, Cylindro-
teuthis magnificns (d'Orbigny), Lower Portlandian. c, Belemnitella
mucronata d'Orbigny, Campanian (Upper Chalk), showing alveolar
slit and groovings of surface, d, Actinocamax verns Miller, Cam-
panian (Upper Chalk). Dotted lines inddcate the destroyed part of the
guard, e, Duvalia lata (Blainville), Neocomian. Some species of
Duvalia are more compressed than this (e'). f, Dactyloteuthis acuarius
(Schlotheim), Charmouthian. Longitudinal section, showing sudden
change of shape and cavity (black), a-c, Ventral views ; d,ft sections ;
e, side view ; a'-e', cross-sections. (All X id.) (a, b, After Pavlow ;
c, e,f, after d'Orbigny ; d, after Crick.)

which the phragmocone is a more decided cyrtocone
than in any belemnite : this leads to the recent Spirula,
a gyrocone in which all trace of guard has disappeared.

THE CEPHALOPODA 189

Alongside the typical Belemnoids of the Mesozoic runs
an allied stock (Phvagmoteuthis, Trias ; Belemnoteuthis,
Upper Jurassic ; Conotetithis, Lower Cretaceous), in which
the guard is a very thin investment of the phragmocone.
By reduction of the phragmocone and guard with re-
tention of the proostracum, this may have led to the
modern squid, Loligo.

Short Bibliography.

BARRANDE, J. — Systeme Silurien de Boheme: vol. ii.,
Cephalopodes (1865-70).

BUCKMAN, S. S. — (i) Yorkshire Type Ammonites,
2 vols. (1909-19), continued as Type Ammonites (1919- ) ;

(2) Pal. Soc., see below; (3) numerous papers in Quart.
Joiirn. Geol. Soc. (1881-1918).

CLARKE, J. M. — The Protoconch of Orthoceras, American
Geologist, vol. xii. (1893).

CRICK, G. C. — (i) Muscular Attachment of Animal to
Shell in Ammonoidea, Trans. Linna'an Soc., vol. vii.
(1898) ; (2) On the Aperture of a Baculite, Proc. Malac.
Soc., vol. ii. (1896); (3) On the Proostracum of a Belem-
nite, the same ; (4) numerous papers in the same Proc.,
and in Geol. Mag., etc.

D'ORBIGNY, A. A. — Paleontologie Fransaise. Terrains
Jurassiques, Cephalopodes, 2 vols. (1842-49); Terrains
Cretaces, Ceph. (1840).

DOUVILLE, H. — Ceratites de la Craie, Bull. Soc. Geol
France, vol. xviii. (1890); and many other papers in the
same periodical.

HYATT, A. S. — (i) Genesis of Arietidae, Smithsonian
Misc. Coll., No. 673 (1889); (2) Pseudoceratites of the
Cretaceous, U.S. Geol. Surv., Monograph 44 (1903);

(3) various papers in Proc. Boston Soc, Nat. Hist, (from
1875).

ipo PALEONTOLOGY

HYATT, A. S., AND SMITH, J. PERRIN. — Triassic Cepha-
lopoda of America, U.S. Geol. Surv., Prof. Paper 40 (1905).

LISSAJOUS, M. — Quelques Remarques sur les Belem-
nites Jurassiques, Bull. Soc. Hist. Nat. Mdcon (1915).

MOJSISOVICS, E. — A series of monographs on Triassic
Cephalopoda published by the various learned societies
of Vienna. For the chief see Geol. Mag. (1908), p. 189.

NEUMAYR, M. — Ueber unvermittelt auftreteuden
Cephalopodentypen, Jahrbuch Geol. Reichsanstalt Wien.,
vol. xxviii. (1878), and other papers published by the
same body.

QUENSTEDT, F. A. — Ammoniten des Schwabischen
Jura (Stuttgart, 1885-88).

SMITH, J. PERRIN. — Carboniferous Ammonoids of
America, U.S. Geol. Surv. Monograph 42 (1903).

SPATH, L. F. — Development of Tragophylloceras loscombi,
Quart. Journ. Geol. Soc., vol. Ixx. (1914).

TRUEMAN, A. E. — Evolution of the Liparoceratidae,
Quart. Journ. Geol. Soc., vol. Ixxv. (1919).

PAL^EONTOGRAPHICAL SOCIETY'S MONOGRAPHS.

BUCKMAN, S. S. — Inferior Oolite Ammonites, Part I.
(1887-1907).

FOORD, A. H. — Carboniferous Cephalopoda of Ireland
(1897-1903).

PHILLIPS, J. — British Belemnites (1865-1909). In-
complete,

SHARPE, D. — Fossil Mollusca of Chalk of England :
Part I., Cephalopoda (1853).

WRIGHT, T. — Lias Ammonites (1878-86).

BRITISH MUSEUM CATALOGUES.

FOORD, A. H. — Fossil Cephalopoda, vols. i. and ii.
(Nautiloidea) (1888-91).

FOORD, A. H., AND CRICK, G. C. — Vol. iii. (Goniatites)
(1897).

THE TRILOBITA AND OTHER
ARTHROPODA

i. Calymene blumenbachi (Fig. 55), the "Dudley
locust," is a familiar fossil, beautifully preserved in the
Silurian Limestone of Dudley in the Black Country,
formerly much worked as a flux for the iron-furnaces,
but now worked out, This was evidently an animal of
very different construction from those we have hitherto
considered. The main part of its body is seen to be
composed of a row of very short, broad divisions (somites,
or segments), thirteen in number, all alike except for a
slight diminution in size backwards. In front is a semi-
circular headpiece (cephalon), which shows suggestions of
somites like those behind ; and at the other end is a
tail-piece (pygidium) showing very well-marked segmenta-
tion (division into somites). From end to end run two
nearly parallel grooves, which divide the whole body into
a median strongly-arched portion (mesotergum) and lateral
flatter portions (pleural regions). The existence of these
three portions of the body suggested the name TriloUta
for the group of fossils to which Calymene belongs.

The trilobites are quite extinct and unknown outside
the Palaeozoic system. There are, however, abundant

191

19* PAL/EONTOLOGY

living forms with so similar a general structure of
skeleton that we can safely refer them to the same
phylum. This is the great phylum Arthropoda, which in-
cludes the Insects, the Myriapods (centipedes and milli-
pedes), the Arachnids (scorpions and spiders), the
Crustacea (lobsters and crabs), and one or two minor
and unfamiliar classes. In all these, the body is meta-
merically segmented (divided into a longitudinal series of
more or less similar somites), and encased in an exo-
skeleton of chitin which may be hardened by a deposit of
calcium phosphate ; a number of the somites at the
anterior end are more or less completely united into a
head ; and on the ventral surface is a paired series of
jointed limbs (usually one pair to each segment) which
fulfil the functions of locomotion (walking or swimming),
the seizing of food, and in aquatic forms respiration.

Shell-growth in the Arthropoda is totally unlike that of
Brachiopoda and Mollusca. The exo-skeleton is not
secreted by a mantle but by the whole surface of a com-
plex jointed body. It is a continuous cuticle, thickened
and hardened to form a number of rigid pieces (sclerites)
but remaining thin between, where flexibility is neces-
sary. Marginal growth is therefore impossible. The
growing animal at intervals bursts and casts off its
whole exo-skeleton, and presently secretes a new and
larger one : this is the process of ecdysis or moulting. The
only way in which the ontogeny of a fossil arthropod can
be determined is by the finding of a series of cast skins
(or of prematurely dead animals) : this has been success-
fully done for a number of species of trilobites.

THE TRILOBITA AND OTHER ARTHROPODA 193

The limbs or appendages of Arthropods are usually of
the greatest importance for classification. Unfortu-
nately the limbs of trilobites are very rarely found.
The ventral exo-skeleton of aquatic Arthropods has
always much more thin cuticle and smaller sclerites
than the dorsal, but in trilobites it seems to have been
nearly all thin cuticle, which easily decayed and allowed
the limbs, themselves soft, to be lost.

Calymene blumenbacki is sometimes found lying flat,
sometimes rolled-up exactly in the way in which the
modern wood-louse rolls itself up when alarmed — that is
to say, with the ventral sides of head-piece and tail-piece
in close contact (compare Fig. 59, e). Comparison of
specimens in the two conditions shows that neither
head nor pygidium is flexible, but that the somites of
the intermediate region (commonly called the thorax) are
movable on one another by pivots at the sides, their
dorsal portion sliding partly over one another when the
animal stretched itself out, and becoming fully exposed
when it rolled up.

In an ordinary-sized specimen 50 mm. long, the head
measures about 16 mm. in length, the thorax 27 mm.,
and the pygidium 7 mm. Subsequent measurements
refer to such an one, but the species sometimes attained
nearly double these dimensions.

Each thoracic or free somite consists of a central arched
portion (axis) and a pair of lateral pleura. The axis of
the first somite is about 1 1 mm. wide, that of the last
about 8 mm. : each consists of a prominent arch, measur-
ing 2 mm. from front to rear, and an anterior sunken

13

I94 PAL/EON TOLOGY

area, of which about 1-5 mm. is visible in an enrolled
specimen, but which almost disappears in a straight one.
The pleura vary in width from about 12 mm. in the first
somite to 10 mm. in the last ; about 4 mm. from the axis
each bears a prominent little forward projection, fitting into
a socket in the somite in front, and serving as a pivot. In
enrolled specimens the part of the pleuron external to
this pivot is largely concealed beneath the somite in front.
A well-marked groove divides each pleuron into a smaller
anterior and larger posterior portion.

The head shows the same trilobed character as the
thorax : the central part is called the glabella, the lateral,
the cheeks or gents. Evidences of segmentation are
shown chiefly by the glabella ; there is, however, one
well-marked cross-furrow (neck or occipital furrow) near
the posterior margin, marking off both in glabella and
genae a posterior somite which closely resembles a
thoracic somite, but is immovably fixed to the rest of the
head-shield. The other signs of segmentation are three
pairs of grooves (segmental furrows] in the glabella, more
or less transversely placed, but not crossing the centre
of the glabella, which is quite smooth. The most anterior
of these is very slight, the second longer or deeper, the
third still longer and deeper. The two latter bend
obliquely backward, so as to mark off rounded lobes
joined to the rest of the glabella by a constricted base.
These three furrows and the occipital furrow indicate
that the head is composed of five united somites With
the disappearance of flexibility in the head, why have
traces of the grooves between the somites been retained ?

THE TRILOBITA AND OTHER ARTHROPODA 195

The reason can be understood by examining the
methods of muscular attachment in Arthropoda. Where
powerful muscular action is necessary the inner face of
the exo-skeleton does not provide a sufficient area of
attachment, and to increase that area a portion of the
cuticle becomes, as it were, pushed in to the interior of
the body. Such an inward process is called an apodeme,

rs.

FIG. 55.— CALYMENE BLUMENBACHI, BRONGNIART, WENLOCK

LIMESTONE.
(Natural size.) (Original.)

F.S., Facial suture ; Gl.%, third lateral lobe of glabella ; O.F., occipital
furrow; PI., pleuron.

and in the thorax of trilobites apodemes are formed in
pairs at the posterior margin of each somite. The
lateral furrows of the glabella occur in corresponding
positions and are evidently apodemes of the head. Some
trilobites have lateral furrows much more like those of
the thorax than in the case of Calymene ; in others they
diverge still more from the primitive characters. Since

196 PALEONTOLOGY

apodemcs serve principally for the attachment of the
limb-muscles, this seems to indicate that in Calymene the
head-limbs are more specialized than in some trilobites,
but less so than others. Of the nature of the head-limbs
(jaws, etc.) of trilobites we know very little.

The glabella of Calymene becomes gradually narrower
forwards. It ends bluntly, and in front of its end there
is a raised border, which margins the whole head-shield,
but is slightly arched up in front of the glabella.

The genae within this raised border are only slightly
convex. Each carries a large eye, opposite the first
rounded glabella-lobe (third head somite). The eyes are
of the compound type, found only in arthropods : instead
of one adjustable lens there are many, of fixed curvature,
each focusing a small segment of the field of vision on
a retinula. The compound character is not well seen in
Calymene, as the surface of the eye is smooth.

On each cheek there is seen a fine dividing line (facial
suture], which starts at the genal angle (outer end of
posterior margin of head), runs first obliquely forward,
and then transversely, until it reaches the posterior
margin of the eye ; here it curves round the inner
margin of the eye, and continues straight forward to the
front margin of the head-shield, where it continues on
the under-side. Just below the margin a transverse
'suture joins the right and left facial sutures.

The most probable explanation of the facial suture is
that it is a line of easy separation to facilitate the moult-
ing (ecdysis) of the head-shield, and especially of the eye,
the lenses of which are formed from the cuticle and must

THE TRILOBITA AND OTHER ARTHROPODA 197

be moulted. The portion of each cheek lying external
to the suture is called the free cheek, the remainder is the
•fixed cheek, and the glabella together with its attached
fixed cheeks constitutes the cranidinm. Loose cranidia
and free cheeks are often found, probably detached in
the process of moulting. In all trilobites which possess
both eyes and facial suture, the eye lies on the free
cheek in contact with the suture, and the adjacent part
of the fixed cheek has a protuberance fitting into the
outline of the eye, called the palpebral lobe.

If the suture cuts the lateral margin in front of the
genal angle, it is described as proparian ; if shifted so
as to cut the posterior margin, it would be opistho-
parian. That of Calymene is on the verge between the
two.

In specimens which show the under-side of the head,
there is seen, articulated to the central part of the
marginal rim, a somewhat oblong plate with indented
posterior margin. This is called the labrum (or hypostome),
and formed a biting upper-lip to the mouth (Fig. 61).

The pygidium of Calymene is nearly semi-circular, the
free posterior margin having much less curvature than
the anterior margin which articulates with the thirteenth
thoracic somite. The axis tapers backwards to a rounded
termination, a little short of the actual end of the pygi-
dium : its anterior two-thirds are divided into six well- -
marked somites, the remainder is unsegmented. There
are six pairs of pleura, which from front to rear show an
increasing backward curvature, until the last pair run
straight back parallel to one another. The margin is

198 PALEONTOLOGY

entire. The pygidial pleura are covered with little
tubercles.

Very little is known of the limbs of Calymene : traces
of them have been found in sections of enrolled specimens,
including spiral structures which are probably gills
carried on the limbs.

2. Dalmanites caudatus (often called Phacops cau-
datus) is another famous Dudley trilobite (Fig. 56). The
most important differences between it and Calymene are
these : The marginal rim of the head-shield is drawn out
into a slight point in front and into a pair of genal spines
behind. The glabella is much wider in front, instead of
narrower ; its lateral furrows are nearly straight and do
not mark off rounded lobes as in Calymene, but converge
towards the middle line, leaving a much narrower smooth
central area. The eyes are very large, usually well-
preserved, and the compound character is obvious, the
surface being divided into a large number of corneal
facets arranged in a very regular manner. The facial
suture is very distinctly proparian, cutting the lateral
margin some way in front of the genal angle and making
a right-angled bend at the eye ; in front it runs round
the front margin of the glabella from one side to the
other, not cutting the front margin of the head-shield,
so that the two free cheeks form one inseparable piece.
The labrum is somewhat triangular (Fig. 61, d):

The thorax consists of eleven somites, which differ
from those of Calymene chiefly in the relative narrow-
ness of the axis (about one-quarter the total width,
instead of nearly one-third).

THE TRILOBITA AND OTHER ARTHROPODA 199

The pygidium consists of twelve or more somites, of
which only ten have pleura recognizable : it is longer
proportionately than that of Calymene, more triangular in
outline, its axis narrower, and at its posterior end it is
drawn out to a long spine (mucronait). The presence
of this spine shows, by analogy with living Arthropods,
that it obtained its food by thrusting its head forward

FIG. 56. — DALMANITES CAUDATUS (BRUNNICH), WENLOCK

LIMESTONE.
(Natural size.) (After Salter.)

F.S., Facial suture.

into the mud, using the tail-spine as its support. The
highly-developed eyes, however, show that it did not
live buried in the mud : the eyes could have been no use
in the search for food ; they must have served to warn
it of the approach of enemies. Their position, on the
highest part of the head, is clear evidence that this was a
bottom-living or benthic animal : in trilobites of Hectic

200 PAL/EON TOLOGY

(swimming) habit, the eyes are close to the margin
of the head, so that they could look outwards as well as
upwards, and in these there is never a tail-spine. We
owe these observations to Prof. Dollo of Brussels.
Nothing is known of the limbs of Dalmanites.

3. Triarthrus becki (Fig. 57) is a small trilobite of
which very beautifully preserved specimens have been
found in the Ordovician Utica Shale of Rome, New York
State. In form this is much less tapering than the pre-
vious forms, and ends more bluntly behind. The semi-
circular head-shield has a much less distinct marginal rim,
the glabella is of almost uniform width, with lateral furrows
which are but very slightly inflected from the transverse
direction. In all these points the head shows much less
difference from the thorax than that of Calymene or
Dalmanites, so that Tnarthvns is a more primitive form.
The cheeks are each little more than half the width
of the glabella ; a long and narrow eye is in the centre
of each ; the facial sutures are opisthoparian, but cut the
hind margin very close to the genal angle, running
obliquely inward and forward (with a bend at the eyes)
to cut the anterior margin separately at a distance apart
not very much less than the width of the glabella.

The thorax consists of fourteen free somites, of which
the first eight are of almost uniform width, after which a
slight tapering takes place: each bears a median
tubercle, as also does the last (occipital) somite of the
head and the first of the pygidium. The pygidium is
very short (about two-fifths as long as broad), and consists
of six somites fused.

THE TRILOBITA AND OTHER ARTHROPODA 201

The labrum is parabolic in outline. From beside it
springs a pair of long, unbranched antennae, which are
the appendages of the first head-somite, and the only
pair not conforming to the general type of the rest.

FIG. 57. — TRIARTHRUS BECKI, GREEN, ORDOVICIAN (UTICA

SHALE), ROME (NEW YORK). (x2.)

(After Beecher, modified.)

The lower figure is a diagrammatic cross-section : on the left side the
exopodite with its fringe is alone drawn, on the right the exopodite
(without the fringe) and endopodite are shown. (NOTE. — The abbre-
viations Ex. and End. have been accidentally interchanged.) An.,
antenna; End., exopodite; Ex., endopodite; F.S., facial suture;
Gn., gnathobase ; M. Ap., mouth appendages ; Fl. , pleuron ; Th. Ap.,
thoracic appendages.

Any one of the largest appendages (from the first
eight thoracic somites) shows the following characters :
It is attached to the ventral surface of the body by
a joint called the coxopodite, from which a hard jaw-

202 PALEONTOLOGY

process (gnathobase) projects towards the middle line.
In the outward direction two branches spring from the
coxopodite (hence the appendages are said to be biramous) :
the inner one, or endopodite, consists of five movable
joints, not unlike one of the ordinary legs of a crab ; the
outer, exopodite, consists of one long joint followed by a
great number of very short joints, the whole fringed with
numerous hair-like setce. Evidently the endopodite is
adapted for crawling on the bottom, the exopodite for
swimming through the water. The function of the
gnathobase is known from the habits of modern
Crustacea with very similar appendages. The right and
left gnathobases in any somite approach one another to
grasp any solid body between and pass it forward to the
pair in front, and so on until it reaches the mouth, where
it is swallowed or rejected according to its being edible
or not.

Thus the appendages of Triarthrus perform the three
functions of crawling, swimming, and seizing food, as do
those of such a Crustacean as Apns : to these may be
added probably that of respiration, for they present a
large surface of thin cuticle to the sea- water in a region
where it must be constantly agitated by the movements
of the limbs. In more specialized trilobites we may
expect a greater division of labour — some appendages
will be specialized as jaws, others for crawling, and
others for swimming (unless one of these two means of
locomotion is lost altogether), and definite gills will
appear as outgrowths from some appendages instead of
the whole surface being used for respiration. In Calymene

THE TRILOBITA AND OTHER ARTHROPODA 203

there is evidence of such gills. The only indication of
specialization in Triarthrus is that in the four pairs of
head-appendages (behind the antennae) the gnathobases
are larger, and the exopodite and endopodite smaller, so
that (as might be expected in the neighbourhood of the
mouth) the jaw-function is beginning to predominate
over the other functions.

4. Trinucleus concentricus (Fig. 58) is a trilobite
found in the Upper Ordovician Shales of Shropshire
and Wales. Complete specimens are rarely found, but
the usual (though not maximum) length appears to be

FIG. 58. — TRINUCLEUS CONCENTRICUS EATON, CARADOCIAN.
(Natural size.) (After Phillips.)

about 17 mm., of which 9 -5, 4-5 and 3 mm., measure head,
thorax and pygidium respectively. The greatest width
of the thorax is 12 mm., that of the head 18 mm. This
great difference is due to the presence of a remarkable
flat brim projecting from the free margin of the head, and
perforated by a number of pits, arranged in four or five
roughly concentric rows. This brim is about 3 mm.
wide at each genal angle, where it contracts into a long,
outwardly-curved genal spine, which extends back 4 or
5 mm. beyond the end of the pygidium ; at the front 'end
of the head the brim is about 2 mm. wide, so that the
length of the head without the brim is about equal to that

204 PALEONTOLOGY

of the thorax and pygidiurn. Consequently in an enrolled
specimen only the head is seen on one surface, and only
the thorax and pygidium with the projecting brim and
genal spines on the other : in fact it is rather doubled-up
than rolled-up.

The head within the brim consists of a highly raised
glabella widening forwards from about 3 mm. to about
5 mm., and smooth protuberant cheeks. The glabella
shows scarcely any sign of segmentation, and the cheeks
show no trace of eyes or facial suture. Trinucleus is thus
a blind trilobite, and has probably arrived at that con-
dition by degeneration consequent on complete adaptation
to the mud-grubbing life for which its shape is so well
suited. In certain species of Trinucleus (or closely allied
genera) there are traces of eyes in the young, and the
late Cambrian genus Orometopus, probably the ancestor
of Trinucleus, has well-developed eyes and opisthoparian
suture. With the loss of the eyes moulting became an
easier process and a facial suture could be dispensed
with, splitting taking place along the margin. The dis-
appearance of glabellar furrows is also a specialization,
and traces of them can be seen in some species of
Trinucleus (as T.favus).

The thorax consists of six somites ; the axis is only
3 mm. wide, or one-fourth the total width. The pygidium
is short, obtusely triangular in shape, the axis tapering,
the number of somites about four.

From what has been said in the chapter on Lamelli-
branchia, it will be understood that in seeking a natural
classification of trilobites it is essential to distinguish

THE TRILOBITA AND OTHER ARTHROPODA 205

between (i) differential or static characters that keep
fairly constant through one line of descent; (2) progressive
characters that are likely to show similar sequences in
several lines of descent ; and (3) adaptative characters
that may occur over and over again in the same or
different branches. Let us first distinguish the progres-
sive characters.

Arthropods are doubtless descended from the meta-
merically - segmented worms (annelids), and have
advanced beyond them by, among other things, the
coalescence of a number of anterior somites (six in
trilobites) to form a head : this process has been termed
cephalization. Among trilobites only, a similar process
takes place at the posterior end, somites fusing to form
the pygidium (caudaltzation), so that we have forms with
very small pygidia consisting of the unsegmented end of
the body alone or with addition of a very few somites
(micvopygons stage), forms with a pygidium including
more somites but still much smaller than the head
(heteropygous), and forms with a pygidium as large as the
head (isopygous). These stages were passed through
independently by various lines of descent, the number of
thoracic somites diminishing as the pygidium incorporated
more of them.

These changes do not seem to bear any relation to
habitat or mode of life, but there are others that do
(adaptative characters). The ordinary trilobites seem to
have lived on the sea-bottom, but to have been able both
to crawl and to swim ; but particular genera or families
took, some to a mud-grubbing, others to a surface-

206 PAL/EONTOLOGY

swimming life, and became adapted to those special con-
ditions. Adaptation to life in the mud is shown by a
shovel-like head ; the eyes as far from the margin as
possible (to be out of the mud, where they are useless)
and sometimes lifted up on long stalks, or else the eyes
are lost altogether; the pygidium sometimes ends in
a spine. Adaptation to a pelagic life is shown by a more
or less globular head with eyes near the margin, some-
times by very thin shell, peculiarly shaped pygidium, and
great development of spines (as a protection).

There remain the static or differential characters, the
chief of which is the course of the facial suture. Others
of less value, because liable to be modified by progression
or adaptation, are the form and proportions of glabella
and cheeks, and of axis and pleura.

The recognition of the importance of the facial suture
is due to the late Prof. Beecher of Yale, who in 1895
proposed the names Opisthoparia and Propana for the
orders with the two types of suture. Unfortunately a
few trilobites, mostly blind, show neither type, and
in these Beecher thought the facial suture was marginal
or sub-marginal : he therefore united them as Hypopana.
More recent discoveries seem to show that these forms
had lost their sutures through their blindness, and can be
traced back partly to opistho- and partly to pro-parian
ancestors. Only two orders appear necessary therefore.
Of these, Opisthoparia is much the larger, and is divided
by Prof. Swinnerton of Nottingham into sub-orders as
shown below.

THE TRILOBITA AND OTHER ARTHROPODA 207

ORDER I. : OPISTHOPARIA.

Facial suture cutting posterior margin.

SUB-ORDER i. Mesonacida. — Long crescentic eyes
close to glabella, wide free cheeks (in earliest genera,
facial suture not yet developed). Nearly all in micro-
pygous stage ; thoracic pleurae to a large extent free
from one another, the free part (pleural spines) more and
more curved back towards posterior end. Mainly
Lower and Middle Cambrian, including the zonal genera
Olenellus (Figs. 59, a; 61, a) and Paradoxides ; the special-
ized pelagic Remopleurides is Ordovician.

SUB-ORDER 2. Conocoryphida. — Glabella' usually
parallel - sided or narrowing forwards, eyes small ;
thoracic pleurae not spiny ; pygidium in most cases
hetero- or iso-pygous. A very large group, ranging
through the whole Palaeozoic era, and subdivisible into
sections :

(i.) Conocoryphina : specialized in respect of blindness,
but otherwise the most primitive ; glabella with oblique
segmental furrows ; free checks extremely narrow ;
micro- to hetero-pygous. Cambrian.

(ii.) Olenina, including two main families, the earlier
Oknida with broad short head, narrow glabella, tapering
thorax of twelve to twenty-two somites, micro- to hetero-
pygous, Cambrian (Fig. 59, c) ; and the later Proetida with
longer head, thorax of eight to ten somites, not tapering,
hetero- to iso-pygous (Ord.-Perm.) — the only family
surviving the Devonian (Fig. 59, g).

(iii.) Ptychoparina, including Cambrian heteropygous
families near to Conocoryphidae but with eyes, and two
later isopygous families — Asaphida (Figs. 59, b; 6i,b, c\
with eight thoracic somites (Camb.-Ord.), and Illanida
(Fig. 59, d, e) with ten (Ord.-Sil.).

(iv.) Calymenina, a small group in which the facial

208

PAL/EOXTOLOGY

FIG 59. — (For description see p. 209.]

THE TRILOBITA AND OTHER ARTHROPODA 20$

suture cuts the genal angle. It includes two important
genera (both Ord.-Dev.) — the heteropygous Calymene,
already described, and the isopygous Homalonotus, with
several sub-genera, of which Trimerus (Fig. 59, /), with
triangular head and pygidium, is the best known.

SUB-ORDER 3. Trinucleida. — A somewhat divergent
series, of which the more generalized ancestral forms
are not known. The Harpedidce (Ord.-Dev.) are micro-
pygous, with many thoracic somites, but specialized in
having a brim like that of Trinucleus, except that it
extends far back beyond the genal angles. Ampyx re-
sembles Trinucleus in thorax and pygidium, but has no
head-brim, has very narrow free cheeks, and a long
median spine in front. Trinucleiis (Ord.) has been de-
scribed and its eye-bearing ancestor Orometopus (Upper
Camb.) referred to already. Possibly to be placed here
are two small isopygous forms — Skumardia (Upper
Camb.) and Mglina (Ord.), the latter a pelagic form with
enormous eyes (Fig. 59, h), probably a nocturnal animal
descending to the depths of the sea by day and feeding at
the surface at night.

SUB-ORDER 4. Odontopleurida. — Glabella sub-
divided into lobes bearing little relation to the original
segmentation ; free cheeks wide; micropygous, pygidium
with pleura projecting so as to make the margin toothed

FIG. 59. — TRILOBITES.

a, Olenellus thompsoni (Hall), Lower Cambrian. Parker's Quarry,
Georgia, Vermont. (Xj.) (After Walcott.) bt Ogygiocaris [Ogygia]
buchi (Brongniart), Llandeilian. (X^. ) (After Salter.) c, Olenus
cataractesSa\tev, Upper Cambrian. (Xj.) (After Salter.) d, Illcenus
davisii Salter, Ordovician. (X^.) (After Salter.) e, Bumastus
barriensis (Murchison), Silurian. Side view of an enrolled specimen.
(X|.) (After Salter.) /, Trimerus {Homalonotus} delphinocephalus
Green, Silurian. (X|.) g, Pfiillipsia derbiensis (Martin), Lower
Carboniferous. (Xi.) (After H. Woodward.) h, sEglina binodosa
Salter. Ordovician. (Natural size.) (After Salter.) i, Agnostus
princeps Salter, Upper Cambrian . (Xil.) (After Salter. )

2 io PALEONTOLOGY

or lobed. Includes two pelagic families (both Ord.-
Dev.) — Lichadidce, with very thin crust, and Acidaspidce,
very spiny. The rather isolated genus Bronteus (Ord.-
Dev.) shows some relations with these, but it is isopygous,
and its pygidium has an entire margin and the axis is
very short.

ORDER II.: PROPARIA.

Facial suture cutting lateral margin. The most
primitive genus is the Middle Cambrian Burlingia, which
is micropygous, but almost as broad as long. Asso-
ciated with it is the small isopygous Pagetia, which

V
FIG. 60.— CHEIRURUS BIMUCRONATUS (MURCHISON), WENLOCK

LIMESTONE.
(Natural size.) (Original.)

retains the facial. suture lost in the commoner forms
Micvodiscus (Camb.) and Agnostus (Fig. 59, *', Camb.-Ord.).
These three genera have the smallest numbers of free
somites of any trilobites — 2, 4 and 2 respectively.

The more familiar Proparia are only known from the
Ordovician onwards : they have nearly always eleven free
somites.

The Encvinuvidce have very narrow free cheeks with an
almost straight facial suture, and small eyes, sometimes
on long stalks ; a long pygidium with many pleurae, and
often a still larger number of divisions of the axis. The
Cheirurida; (Fig. 60) have a facial suture, with right-angled

THE TRILOBITA AND OTHER ARTHROPODA 211

bend at the eye ; pleura of free somites spiny ; pygidium
small, of three to five somites, with pleura projecting
from margin. This family includes several remarkable
pelagic forms — Spharexochus (Ord.-Sil.), Staurocephalus
(Ord.), Deiphon (Sil.). The Phacopidce are remarkable
for their highly-developed eyes, with many well-pre-
served lenses ; the glabella is always expanded in front ;
the facial suture as in the last family ; pygidium with
entire margin. Dalmanites (Ord.-Sil.) has been described ;
Phacops (Sil.-Dev.) differs in the more expanded

a V^~/ b ••" ^y "> c

FIG. 61. — HYPOSTOMES (LABRA) OF VARIOUS TRILOBITES.

a, Elliptocephala asaphoides Emmons, Lower Cambrian. (X3-) (After
Walcott.) b, Ogygiocaris buchi (Brongniart), Llandeilian. (Xg.)
Dotted lines indicate outline of doublure. (After Salter. ) c, Asaphus
tyrannus Murchison, Llandeilian. (X|.) (After Salter.) d, Dalma-
nites caudatus (Briinnich), Salopian. (X |.) (After Salter.)

glabella, on which the segmental furrows are obsolete,
and the shorter, non-mucronate pygidium.

General Geological History of the Trilobites.

They appear suddenly in the Lower Cambrian, with
every appearance of having been long in existence
already, since specialized forms like Agnostus appear
alongside the primitive Mesonacidte. Most of the Cambrian
forms, however, are micropygous and unspecialized. In
North America different genera and species of Mesona-
cidtz serve as zone-fossils in the Lower Cambrian
(Nevadia, Elliptocepkala, Callavia, Olenellus), and in
Britain and Scandinavia species of Paradoxides serve

212 PALEONTOLOGY

similarly for the Middle Cambrian, and species of Olenidce
for the Upper Cambrian (see Appendix I.). In the
succeeding Ordovician system the trilobites attain their
acme. Although with the modern narrowing of the
extent of the term " family " it is no longer possible to
say that all families of trilobites occur in the Ordovician,
yet there are few that do not, and there is a greater
abundance and variety of forms than in any other system.

Two families are especially characteristic of the Ordo-
vician— Trinucleidtz and Asaphida. Both are strictly
confined to the Ordovician, if we include the Tremadoc
beds, which form a transition from Cambrian to Ordo-
vician, and if we confine ourselves to Britain and Scan-
dinavia ; but asaphids have been found as low as
Middle Cambrian in British Columbia. They thus
afford an instance of migration as an explanation of
cryptogenetic types. Asaphids are more character-
istic of the Lower and Middle Ordovician, trinucleids
of the Upper. Neither family survives the period, but
allies of each persist into the Silurian — Illcenus repre-
senting the asaphids and Ampyx the trinucleids.

In the Silurian the trilobites are still abundant, but no
new families appear and many genera are extinct,
especially among Opisthoparia, the Proparia now in-
creasing in importance. In the Devonian the list of
families is diminished by three — the two mentioned above
as surviving into the Silurian and the Encrinuridce. Lastly
in the Carboniferous only one family survives, the
Proetidce, represented by four genera, of which one
survives into the Permian of tropical regions.

THE TRILOBITA AND OTHER ARTHROPODA 213

Besides their value as indices of age, trilobites may
serve another purpose to the geologist. The majority
being bottom-living forms (benthic) they were more re-
stricted than pelagic forms in geographical extension.
They may therefore assist in the delimitation of marine
zoological provinces in Palaeozoic times, and so help
towards the reconstruction of past geographies. Such
provinces are regarded by some geologists as established
for the Cambrian, Ordovician, and Devonian periods,
though the Silurian gives little indication of provinces.
Thus, in the Cambrian of China there are trilobites
belonging to genera unknown in Europe, and again in
the Rocky Mountains there is another series of forms,
while in Newfoundland the European faunas are found.
There is therefore a probability of separate provinces
between which migration was difficult/* At the same
time great caution is necessary in drawing conclusions
from limited data. Evidently if two faunas from different
parts of the world contain no species in common, we
cannot be sure that they were contemporaneous : it
may be that they belong to different zones in the system,
and that one zone was never deposited in the one area,
or was subsequently removed by denudation, while the
same was the case with the other zone in the other area.
Mistakes of this kind have been made and afterwards
corrected in several cases, and it may be that the same
will happen to the supposed Palaeozoic life-provinces.

As more detailed knowledge is obtained of the strati-

i

* See Cowper Reed, " Pre-Carboniferous Life Provinces.'
Records Geol. Surv. India, vol. xl. (1910).

214 PAL/EONTOLOGY

graphy of little-known areas like China and British
Columbia, it will be possible to say with greater certainty
whether they were inhabited by faunas differing from
the strictly contemporaneous faunas of Europe.

Ontogeny of Trilobites.— Owing to the inability of
the exo-skeleton to grow with the animal's growth, the
adult trilobite, unlike a brachiopod or mollusc, retains
no trace of its early form. The only way to trace the
development of a trilobite is to find remains of a sufficient
number of individuals of all stages of development
(whether cases of infant mortality, or merely the skins
moulted in healthy development) to be sure they belong
to the same species. This has been done in a number
of cases, particularly by Barrande in Bohemia and
Beecher in North America, and the general course
of development has been established. The earliest
stage known has been termed the protaspis larva, and
is probably the form in which most trilobites were
hatched. At this stage, the whole dorsal shield is
less than i mm. long, and consists principally of head,
the thorax and pygidium being very rudimentary. The
trilobation is distinct, and the glabella shows five well-
marked somites. The eyes are marginal, probably as an
adaptation to the free-swimming life usual in Crustacean
larvae. By repeated moults the larva passes through
stages in which the post-cephalic part of the body
increases in length and in number of somites, gradually
attaining the adult state.

The remaining divisions of the Arthropoda must be
dealt with much more briefly.

THE TRILOBITA AND OTHER ARTHROPODA 215

PHYLUM: ARTHROPODA.

Metamerically segmented animals with jointed appen-
dages, the anterior somites fused into a head.

CLASS : CRUSTACEA.

Usually aquatic, and breathing by gills borne on the
appendages, with one or two pairs of antennae.

SUB-CLASS : Trilobita. — Classification already given.

SUB-CLASS: Entomostraca. — A rather heterogeneous
group of primitive Crustacea, all passing through the
stage of a Nauplius larva, with only three pairs of
appendages. Chief orders — Phyllopoda, Ostracoda,
Cirripedia.

SUB-CLASS: Malacostraca. — With 19 somites (or in
Phyllocarida a few more) the head of 5 (bearing two pairs
of antennae, i of mandibles, 2 of maxillae), the thorax of
8 with walking limbs (of which the first i, 2, or 3 may
be fused with the head). Chief orders — Phyllocarida,
Isopoda, Decapoda (with sub-orders Macrura and
Brachyura).

CLASS: ARACHNIDA.

Without antennae. Somites 18 or fewer.

SUB-CLASS : Merostomata. — Marine, breathing by
gills. Orders- -Xiphosura, Eurypterida.

SUB- CLASS : Arachnida (sensu stricto}. — Terrestrial,
breathing by lungs or tracheae. Chief orders —
Scorpiones, Araneae.

CLASS: MYRIOPODA.

With numerous similar somites. Terrestrial, breath-
ing by tracheae.

2i6 PALEONTOLOGY

CLASS: INSECTA (Hexapoda).

With head bearing antennae, mandibles and two pairs
of maxillae, thorax with 3 pairs of appendages and 2 pairs
of wings, and abdomen without appendages.

Notes on the Chief Classes and Orders of
Arthropoda (other than Trilobites) found Fossil.

The Phyllopoda are mostly non-marine, but as common
in excessively salt as in fresh waters. The form of the
body and the appendages have some resemblance to those
of a primitive trilobite like Olemis ; but (i) there is no
trilobation ; (2) no fusion of posterior somites into a
pygidium ; and (3) the exo-skeleton of thorax and abdomen
is thin, and to protect the hinder regions the head-shield
is extended back as a loose covering. In some forms,
especially in the fossil Estheria, this head-shield is divided
into two by a median suture so as to form practically a
bivalve shell : this being ornamented by sub-concentric
ridges closely resembling the growth-lines of a mollusc,
it is very easy to mistake Estheria for such a lamelli-
branch as Posidonomya. All the more is this the case,
because while, in general, arthropod shells do not grow
marginally but are only replaced by ecdysis, Estheria is
stated to be a unique exception, in which the concentric
lines are actually growth-lines. The following points
should be looked for to distinguish Estheria : (i) There is
no hinge-structure as in lamellibranchs ; (2) the material
of the shell is not calcareous but chitinous; (3) with
a lens there may be seen between the apparent growth-
lines a network of striae which is never seen in lamelli-
branchs. Estheria is common from Carboniferous to
Triassic. Protocaris is a Cambrian marine form scarcely
distinct from the modern freshwater Aptis.

THE TRILOBITA AND OTHER ARTHROPODA 217

The Ostracoda are more thoroughly bivalve than
Estheria, as the two valves completely enclose the whole
body, and they are calcareous. They never show any
" growth-lines," and are nearly all of too small a size to
be lamellibranchs. The Silurian Lepevditia is a giant
among ostracods, being nearly 20 mm. in length : a
glance at the hinge will show that it is not a lamelli-
branch : there is also a tubercle on each valve, antero-
dorsally, which could not occur in a molluscan shell.
Although the majority of ostracods are marine, it is in
freshwater deposits (such as the Purbeck Marls and
Wealden Shales) that they occur in such abundance as to
attract attention in spite of their small size. Ostracods
range from Cambrian to Recent.

The Cirripedia are the only arthropods which are
fixed. They are all marine, and occur sparsely in forma
tions from the Ordovician onwards. They secrete a
shell of five or more pieces, which shows little analogy
with that of any other arthropod ; and in the works of
early palaeontologists such as Cuvier and the Sowerbys
they will be found figured among Mollusca.

The Phyllocarida are an almost extinct group, mainly
Palaeozoic, differing from phyllopods in little else than
the great restriction in the number of somites, and the
lesser backward extension of the head-shield. Some
forms (Hymenocaris, Echinocaris) are fairly common fossils
from Cambrian to Devonian.

The Isopoda (wood-lice) are terrestrial forms, with a
curious external resemblance to trilobites, which some
palaeontologists regard as a real relationship. Unlike
trilobites, however, their number of somites is fixed.
They are rare as fossils, though found occasionally from
the Devonian upwards : one species, Arckaomscus brodiei,
occurs in enormous numbers in certain of the Lower and
Middle Purbeck beds of the South of England.

218 PALEONTOLOGY

The Decapoda Macrura (lobsters, etc.) occur in
various formations from the Triassic upwards. Well-
preserved specimens are abundant in the Lithographic
stone of Solnhofen, Bavaria.

The Decapoda Brachyura (crabs) first appear in the
Middle Jurassic. A species, Pal^ocorystes stokest, abounds
in one bed in the Gault and (as a derived fossil) in the
Cambridge Greensand. Other forms occur in phosphatic
nodules in the London Clay.

The Eurypterida had the body divided into head,
mesosoma, and metasoma, each of six somites. The
head-appendages all bear gnathobases, none taking
the character of antennae (a fundamental distinction
between Arachnida and Crustacea) ; they increase in
size backwards, the last being a very large pair of claws.
There are no walking-limbs behind the head, the appen-
dages of the mesosoma being broad and flat, and bearing
book-like gills, while the metasoma has no appendages.
Behind the eighteenth somite is a telson, which varies
in character from an expanded swimming organ (Ptery-
gottis) to a long and narrow rod (Eiirypterus). In early
Palaeozoic times this order seems to have been confined
to American seas, but towards the end of the Silurian
period it spread over the European area, giving rise to
some gigantic forms (Pterygohts, a nectic form, two metres
long ; Stylonurus, benthic, one metre). Some of the
Devonian and later eurypterids became adapted to a
freshwater habitat, the last of these being found in the
Permian, after which the order was extinct.

The Xiphosura are close allies of the eurypterids,
differing from them in the fusion of the somites of the
meso- and meta-soma into one piece. They range from
Cambrian to Recent, and are mainly marine, but in the
later Palaeozoic some freshwater forms existed. Some
of the earlier genera have a close external resemblance to

THE TRILOBITA AND OTHER ARTHROPODA 219

trilobites, but this is largely the effect of adaptation
to similar conditions of life.

The Scorpiones are direct descendants of the
eurypterids, from which they differ only in the effects
of adaptation to a land-life. They are known from the
Silurian period, and the scorpions of to-day differ very
little from those of that early time.

The Araneae (spiders) are derived from scorpions by
a shortening of the abdominal region. They are known
from the Coal Measures, but are very rare as fossils until
the Oligocene, where they occur in the amber (fossil
resin) of the Baltic area.

The Myriopoda (millipedes and centipedes) are very
rare as fossils, but occur as early as the Devonian.

The Insecta are abundant in occasional beds, of which
the Coal Measures (particularly of Commentry in France)
and the Oligocene amber are the most prolific, but other-
wise very rare. The earliest is recorded from the Upper
Ordovician, and if not misidentified is the only land-
animal known from so early a period.

Short Bibliography.

BEECHER, C. E. — (i) Outline of a natural classification
of Trilobites, Amer. Journ. Sci. (4), vol, iii. (1897).

(2) A series of articles on appendages of Triarthrus and
Trinucleus, Amer. Journ. Sci. and Amer. Geologist, 1893-96.

(3) Larval stages of Trilobites, Amer. Geologist, vol. xvi.

(1895).

DOLLO, L. — La Paleontologie Ethologique, Bull. Soc.

Beige Geol. Pal. Hydr., vol. xxiii. (1909).

RAYMOND, P. E. — Beecher's Classification, Amer. Journ.
Sci., vol. xliii., p, 208 (1917).

REED, F. R. C. — Numerous papers in Quart. Journ.
Geol. Soc. (from 1896) and Geol. Mag. (from 1894).

220 PALEONTOLOGY

SWINNERTON, H. H. — (i) Classification of Trilobites,
Geol. Mag., dec. 6, vol. ii. (1915). (2) Facial Suture of
Trilobites, Geol. Mag., dec. 6, vol. vi. (1919).

VOGDES, A. W. — Classed and Annotated Bibliography
of Palaeozoic Crustacea, California Acad. Sci., Occ. Papers,
iv. (1893), with Supplement in Proc. California Acad.,
vol. v. (1895). Invaluable for all research work on
Trilobites.

WALCOTT, C.D. — Cambrian Geology and Palaeontology,
Smithsonian Misc. Coll., vols. liii. (1910), Ivii. (1914) Ixiv.
(1916), Ixvii. (1918).

VI
THE VERTEBRATA

THE Vertebrata are commonly accepted as the dominant
phylum of the Animal Kingdom. That position might be
disputed by the Arthropoda, for the latter are certainly
more all-pervading and have insinuated themselves
everywhere where life is found, even into many nooks
and crannies which the Vertebrata have left alone. But
it is to the credit of the Vertebrata that they have almost
completely avoided the dirty corners of parasitism and
degeneracy in which arthropods abound, and that their
dominance is mainly a matter of power, of high organiza-
tion, and finally of intelligence.

In the most primitive Vertebrata the chief hard parts
are external, and though an internal skeleton is always
present it was at first of organic material and rarely
preserved fossil ; subsequent evolution is marked in
general by a steady increase in importance of the endo-
skeleton, which becomes a series of articulated bones
hardened by calcium phosphate, and a diminution in that
of the exoskeleton.

The first form taken by the exoskeleton is that still
found in the sharks constituting the " shagreen ": every
point on this rasping surface marks a single "skin-

221

222 PALEONTOLOGY

tooth " or placoid scale, the structure and development
of which are essentially those of teeth (Fig. 64). Our
teeth are in fact the greatly-modified survivors of what
was once a general body-covering, retained on the jaws
for special purposes when no longer needed on the skin.

Water-breathing Vertebrata.

The first British Vertebrata are found near the top of
the Silurian system, in the Ludlow " bone-bed." Here
abound the little skin-teeth of Thelodus (Fig. 63, a),
which, as shown by a complete body (only a few inches
long) found in Scotland, is a primitive member of a
group, Ostracodenni (Fig. 62), the exact taxonomic position
of which is doubtful, as it is not clear whether they
possessed the biting jaws found in all other Vertebrata.
In higher members of the same class, found also in the
Silurian, but more abundantly irnhe Devonian (e.g.,Cepha-
laspis) the skin-teeth became fused into large armour-
plates, and (by adaptation to a similar life) many of them
came to mimic closely the contemporary eurypterids. The
microscopic structure of the armour is quite different in
the two cases, and though much ingenuity has been
wasted in trying to prove an arthropod ancestry for
Vertebrata, we may feel confident that the resemblance
is only a very striking case of convergence. The ostra-
coderms died out at the end of the Devonian.

The undoubted fishes, with typical vertebrate jaws,
have an internal skeleton ; but in the lower of the two
grades into which fishes may be divided (Chondrichthyes),

THE VERTEBRATA 223

this skeleton is composed of cartilage, and is there-
fore rarely preserved fossil. These _are sharks and rays,
known principally to fossil-collectors by their teeth
(Fig. 63), and the bony " ichthyodorulites " which
support the front margin of the fins in some cases.
Ichthyodorulites (Onchus) accompany skin-teeth in the

FIG. 62. — LANARKIA SPINOSA, TRAQUAIR, DOWNTONIAN,

LESMAHAGOW. ( x |.)

(After Traquair.)

This differs from its contemporary, Thelodus, in having spiny skin-teeth.
The head and trunk have been flattened down from above, but the tail
is turned over on its side.

Ludlow bone-bed, Shark's teeth are common in some
beds in most geological systems. They are usually
sharp lacerating teeth, pointing inwards on the jaws,
so as to prevent prey from slipping out ; they are
arranged in rows, new rows developing on the inside
of the jaw throughout life, and moving forwards as the
older teeth fall out. Thus loose teeth are commoner than

224

PALEONTOLOGY

FIG. 63.— FISH TEETH AND SCALES.

a-g, j, teeth ; h, i, scales. The doited part in b-f\s the root ; in h it is the
''"p'irt overlapped by other scales, a, Thelodus pagei (Powrie), Down-
tonian, skin-tooth from above; a', from below, showing nutritive
foramen ; a", from side, b, Hybodus grossiconus Agassiz, Bathonian.
Tooth. (Natural size.) <:, Acrodus nobilis Agassiz, Lower Jurassic.
Tooth. (Natural size.) d, Lamna appendiculata Agassiz, Upper
Cretaceous. Tooth. ( X £. ) e, Corax falcatus Agassi/., Upper
Cretaceous. Tooth. (Natural size.) f, Carcharodon inegalodon
Agassiz, Miocene-Pliocene. Tooth, (xi.) g , Ceratodus altus

THE VERTEBRATA 225

Complete series, and it is difficult to know whether slight
differences among isolated teeth are due to their belong1
ing to different species of sharks or to different parts
of the jaw of the same species.

Like other vertebrate teeth, those of the sharks consist
of a roott buried in the soft tissues, and a crown which
is exposed. Up to the Jurassic period, shark's teeth
have shallow, undivided roots (Fig. 63, b, c) ; from the
Cretaceous period onwards, forms with deep, divided
roots occur (Fig. 63, d, e,f).

The tooth is mainly composed of dentine, a calcareous
tissue full of fine tubules into which pass processes
tfrom the soft tissues of the central pulp-cavity (Fig. 64).
The crown has its surface covered with a thin layer
of enamel, a much denser material than the dentine. In
the fossil state, teeth are usually highly phosphatized,
and may acquire a black colour.

Sharks' teeth are among the most indestructible of
fossils, and worn examples are common among derived
fossils.

In one remarkable Palaeozoic group (Cochliodontidse
and Edestidae) the teeth of the middle line of the jaws
do not fall out but curl in under the front of the jaw
as the new teeth push them forward. In the latest

FIG. 63. — FISH TEETH AND SCALES (Continued).

Agassis, Rhaetic, Tooth,, (xf.) h, Acrolepis semigranulosus
Traquair, Lower Carboniferous. A^ single ganoid scale, X f , surface
pattern omitted. *, Lepidotus minor Agassiz, Purbeck beds (Upper
Jurassic). Part of scaly armour of trunk. (X&) /. Ptychodus
mammillaris Agassiz. Upper Cretaceous. Tooth, surface view;
j', profile. (x£.) The area shown dotted is part of the crown and
should be shown with fine, more or less concentric lines, a, k, After
Traquair; b, c, /, g, after Agassiz ; d, <?, z, /, after Smith Woodward.

15

226 PALAEONTOLOGY

member of this group (Helicoprion, Permo-Carboniferous)
these old teeth are rolled up into a spiral of four or more
whorls.

Some sharks and the rays, instead of swallowing their
prey whole, grind it in their mouths : in these, the teeth
have lost their sharp points and compressed shape, and
form a mosaic with a rough surface, extending over the
palate and part of the floor of the mouth as well as the
jaws (Fig. 63, f).

No attempt can be made here to describe the details
of the endoskeleton of Chondrichthyes, but a few words

n.o.

FIG. 64. — TOOTH STRUCTURE.
(Diagrammatic vertical section.)

d, Dentine; e, enamel; c, crown; p.c., pulp-cavity; n.o., nutritive
opening ; r, root.

may be given to the general form of the body. This is
elongated, tapering towards both ends, laterally com-
pressed in the nectic (swimming) forms, depressed in the
benthic (bottom-living) skates. Besides the two pairs
of fins (answering to the limbs of land-vertebrates) there
are a variable number of median fins, spaced out along
the whole of the dorsal edge, but only on that part of the
ventral edge that lies behind the paired fins. The end
of the tail is bent up and combines with a median
ventral fin to form the " heterocercal caudal fin," charac-
teristic not only of all cartilaginous fishes and some

THE VERTEBRATA 227

bony fishes, but also of such primitive forms as Thelodus
(Figs. 62 ; 65, a).

The higher grade of fishes (Osteichthyes, bony fishes)
have an endoskeleton of bone, but they retain an exo-
skeleton of scales of some sort. They vary greatly
in shape from greatly compressed forms, whose height is
nearly equal to their length, to long cylindrical eel-like
forms. The latter appear to be the senile forms of
a number of families, according to Dr. Smith Woodward ;
the former are, in part at least, highly specialized forms.
Most fishes have a fusiform shape, lying between these
two extremes.

Three divisions may conveniently be recognized among
bony fishes.

i. The Ganoidei. — Though no longer admitted as
a natural group by zoologists, this makes a serviceable
category for palaeontology.

The name refers to the possession of "ganoid " scales
— i.e., large thick enamel- covered scales which typically
articulate together to form a coat of mail, though in some
genera these scales are greatly reduced. The tail is
typically heterocercal, but in some it acquires the super-
ficial symmetry of the " homocercal " type (Fig. 65, b).
Modern ganoids are confined to the Northern Hemi-
sphere and to the fresh waters (except that the sturgeons
migrate between river and sea). Fossil ganoids are found
from the Devonian (Middle Old Red Sandstone) onwards,
and in marine as well as in freshwater deposits. A
familiar example is Lepidotus (Jur.-Cr.et), of which the
scales and teeth are common in the Wealden beds

228 PALEONTOLOGY

(Fig. 63, i). The ganoids were divided by Huxley into
Crossopterygii (fringe-tinned) , in which the paired fins have
a central skeletal axis with a fringe of fin-rays, and
Actinoptevygii, in which the axis is wanting and the fin
supported by rays only.

2. The Teleostei, or typical bony fishes, appear sud-
denly in the Upper Cretaceous. They are distinguished
by their thin, overlapping scales from the typical ganoids,
but some fishes reckoned as ganoids have similar scales.
When preserved as fossils the skeleton is usually fairly

FIG. 65. — CAUDAL FINS.

a, Heterocercal fin of a shark, vertebral axis rising, second lobe of fin
developed below it ; b, homocercal fin of bony fish, vertebral axis as in
a, but the whole fin externally symmetrical ; c, fin of Ichthyosauriis,
externally like b, but vertebial axis running into lower lobe.

complete, and the scales may also be preserved in
position. Isolated vertebrae are sometimes found and
are recognizable by their cylindrical bodies with very
deep conical hollows at each end : this biconcave or
" amphiccelous " type is, however, found also in the
lower grades of nearly all branches of air-breathing
Vertebrata. The neural arches which protect the spinal
cord may remain attached to or be detached from the
bodies ; in the tail there is a similar haemal arch below.
It is impossible to give further details within the neces-
sary limits of this chapter.

THE VERTEBRATA 229

3. The Dipnoi are interesting as being the fishes in
which there is a transition to the internal structure of
air-breathing vertebrates : at least this is the case with
existing survivors of the group, but many extinct forms
are associated with them on account of their skeletal
characters. The Dipnoi appear as early as the ganoids,
and, like them, were both marine and freshwater, but
the three surviving genera are confined to rivers and to
the Southern Hemisphere (in curious contrast to the
ganoids). The best-known of these modern genera is
the Australian Ceratodus (Fig. 66, a), which has a pair
of large ridged palatal teeth, formed by the fusion of
many rows of small teeth. Similar palatal teeth (Fig.
63, g) were known as fossils in the Rhaetic beds before
the living Ceratodus was discovered ; they occur in
Triassic and (very rarely) in Jurassic strata, showing
Ceratodus to have been marine at that time. It dis-
appeared from the Northern Hemisphere like Tngonia,
but somewhat earlier.

Air-breathing Vertebrata.

These are usually divided into Amphibia, Reptiles,
Birds, and Mammals, but palaeontological discoveries
have so extended the idea of a reptile that it is no longer
logical to separate birds from that category, so that only
three classes will be recognized here.

The Amphibia are distinguished by the fact that they
retain signs of their aquatic ancestry by passing through
a larval stage in which they have gills and (with a few

230 PALEONTOLOGY

exceptions) live like fishes : the tadpole-stage of the frog
is a familiar example. . At the present day they are re-
presented only by the primitive and somewhat degenerate
newts and salamanders, the highly-specialized frogs and
toads, and the equally specialized snake-like Coecilians of
the tropics. The two latter orders are unknown before the
Eocene period ; the earliest newt is of Lower Cretaceous
age ; and between them and the other fossil Amphibia
there is at present a great gap in time, no certain remains
of Amphibia being known from anywhere in the
Triassic or Jurassic strata. In the Carboniferous and
Permian systems there are found a series of Amphibia
with complex bony skulls, which resemble those of some
fishes on the one hand and those of reptiles on the other
far more than those of modern newts and frogs. That
they were really Amphibia is indicated by their retention
of gill-arches in the throat, and by the fact that their ribs
do not meet ventrally in a breast-bone, so that they
must have breathed like a frog by swallowing air, not
sucking it in. Unfortunately little is known of the limbs
of these Stegocephalia, and we have no certain knowledge
of the stages through which the fin of the fish, adapted
to serve as a flexible oar, became transformed into a
limb capable of performing the much more complex
movements required by a land-living animal. The type
of fin which is most easily comparable with the terrestrial
limb is that of Ceratodus, found also in some of the
Upper Palaeozoic sharks (Pleur acanthus) and ganoids
(Holopty chins). In this the fin is supported by a cartila-
ginous jointed axis, to which are articulated a series of

THE VERTEBRATA . 231

cartilaginous rays on both margins. If we compare this
with the leg of a newt, which shows the simplest form of
terrestrial limb, we see the essential changes in the latter
are — (i) the lengthening of the axis ; (2) the introduction
in the second joint of a pair of elements, side by side,
which makes possible a movement of rotation impossible
in the fin ; and (3) the great reduction in number of the
rays and their separation to form five fingers or toes,
capable of independent movement (Fig. 66, a, b).

The Reptilia of the present day consist of four or five
small orders — the lizards and snakes, tortoises and
crocodiles being the familiar forms. To these must be
added, as representing a separate order, the New
Zealand lizard Hatteria, sole survivor of an important
Permian and Triassic series. The birds should also be
counted as another order. These are the few scattered
survivors of a series of forms so vast and varied that the
Mesozoic era in which they flourished has been called
the Age of Reptiles. They spread over the surface of
the continents, hitherto almost devoid of animal life, and
exhibited the same " radiative adaptation " to all possible
methods of life that the Mammals showed afterwards in
the Cainozoic era. There were herbivorous reptiles and
carnivorous reptiles, aquatic, terrestrial, and aerial reptiles.
We cannot here deal with the many instructive facts
shown by the study of fossil reptiles, but must confine
ourselves to the case of one order of reptiles, whose
remains are fairly common in some of the Jurassic strata
(Lias), the Ichthyosauria. These occupy the same posi-
tion among Reptiles that the whales do among Mammals.

232

PAL/EONTOLOGY

They underwent complete adaptation to a marine life,
and approximated to fishes in shape and general structure,
but in accordance with the law of irreversibility in
evolution they could not lay aside all the characters they
had gained during their ancestral land-life, nor could they
recover certain fish-structures which their land ancestors
had lost — the gills in particular. If we compare the
limbs (or paddles) of an ichthyosaur with the fin of
Ceratodus on the one hand and the limb of the salamander

N

FIG. 66. — FINS AND LIMBS.

a, Ceratodus ; b, a newt ; c, Ichthyosaurus. The dotted lines connect
homologous structures.

on the other, we see that while in outline, absence of
separate fingers, and slight possibility of movement
between the numerous small, closely -packed bones it
agrees with the fish, its fundamental structure is that of
a land-animal (Fig. 66, c). So, too, with the tail-fin :
externally it resembles the homocercal tail of a bony
fish, but the end of the vertebral column bends down
into the lower lobe (Fig, 65, c). In these and other ways
ichthyosaurs show convergence towards fishes, but the
resemblances are not exact and are mainly external.

THE VERTEBRATA 233

The vertebrae of Ichthyosaurus are common in Jurassic
Clays, and easily recognized by their extreme shortness—-
the length of a vertebral body being only about one
quarter its height or breath ; they are biconcave, and the
arches are loose.

Another Mesozoic order of marine reptiles is the
Sauropterygia (e.g., Plesiosaimis] distinguished by a long
neck and small head (in contrast to the large head and
very short neck of ichthyosaurs), by the vertebrae being
about as long as high, with very shallow concavities at
the ends, and the limbs rather less modified from the
land-type. The earliest crocodiles were also marine, but
the Chelonia (turtles) had their most striking features
developed before they took to the sea.

Along two distinct lines Mesozoic reptiles adapted
themselves to flight. The pterosaurs had a membranous
wing, for the support of which the fifth finger was
enormously lengthened, the leg and tail also helping
to keep it stretched. They lived through the Jurassic
and Cretaceous periods. In the birds, on the other
hand, there are only a short thumb and two fingers, the
latter combining to support the wing but not being
greatly lengthened. (In flying Mammals — bats — yet
another type of membranous wing is developed, all four
fingers being lengthened to support the wing, like the
ribs of an umbrella). The earliest bird (Archteopteryx)
was found in the lithographic stone of Solnhofen, high
in the Jurassic. The few known Mesozoic birds all
have teeth, and Avchaoptevyx alone of all birds has a
long, many-jointed tail.

234 PALEONTOLOGY

The Mammalia are represented in a few Mesozoic
rocks by lower jaws and teeth of very small forms, which
appear to belong to primitive representatives of the two
orders now confined to Australasia, and there represented
by greatly specialized forms — the egg-laying Monotremata
and the pouched Marsnpialia. Why no other remains
should be found than lower jaws has never been satis-
factorily explained. While the Reptiles dominated the
world, the Mammals seem to have been a very lowly
and insignificant group, but on the extinction of most
of the former at the end of the Cretaceous period the
Mammals became the dominant land-animals.

The early Eocene mammals are all of small size, with
the neck not flexible and the trunk passing gradually
into the tail. Their limbs are not greatly removed
in structure from the primitive five-toed (pentadactyle]
type, except for the presence of a projecting elbow in the
fore-limb and heel in the hind-limb : the two bones
in the second limb-segment are free to move on one
another so that the limbs can be rotated. There are
forty-four teeth in all, in continuous series, with low
crowns (brachydont) and a simple arrangement of conical
tubercles on the surface. The brain (as shown by
internal casts of the skull) is small like that of a reptile.

From these primitive forms higher forms are soon
developed in a series of adaptative radiations, like those
shown by the Mesozoic reptiles. All the orders of
Mammalia are thus separated in the course of the
Eocene period. Except for the Insectivora and Rodentia,
which remain small, they all show a gradual increase in

THE VERTEBRATA 235

size, the largest forms being chiefly found in the Pliocene
period, after which glacial conditions exterminated the
majority of them. The brain also increased greatly,
and limbs and teeth underwent various specializations in
accordance with the food and habitat.

The most interesting developments are perhaps shown
by the Ungulata (hoofed mammals). In these the teeth
become high-crowned (hypsodont] to stand long wear,
and develop varying patterns which as they wear down
show a surface partly of hard enamel, and partly of soft
dentine, and therefore never wearing smooth. The limbs
become longer and vertical in position, lifting the body
high off the ground, while at the same time they lost the
power of rotation (useless to a running animal) by the
fusion of the two bones in the second limb-segment. In
one branch of the Ungulata (Perissodactyla, odd-toed)
the middle digit bears the weight of the body, and the
others gradually shorten and finally are neafly lost in the
modern horse (Equus), the ancestry of which has been
very fully worked out. In another branch (Artiodactyla,
even-toed) the weight is borne by the third and fourth
digits (the " cloven " hoof), the first disappearing, as may
also the second and fifth : these include the cattle, deer,
and camels. In the Proboscidea, again, while the teeth
undergo enormous specialization, the Jimbs remain com-
paratively primitive, all five digits persisting, and fusion
of the bones of the second segment taking place in the
fore-limb only. Until recently the Proboscidea were a
cryptogenetic group, appearing suddenly in Europe in
the Miocene period, but their earlier ancestors have

236 PALEONTOLOGY

since been discovered in Eocene and Oligocene strata in
Africa.

An outline classification of Vertebrata is appended
which attempts to combine and harmonize those of
Prof. Goodrich and Dr. Smith Woodward. Brief notes
are added on some of the orders which have not been
already mentioned.

Classification of Vertebrata.

CLASS: CYCLOSTOMATA (without jaws: the
lampreys).

Only one fossil form known : Palaospondylus, Devonian.
(Some authorities associate the Ostracodermi with this
class, instead of with the true fishes.)

CLASS: PISCES.

SUB-CLASS i. Ostracodermi (armour-plated fishes),
Sil.-Dev.

SUB-CLASS* 2. Chondrichthyes (cartilaginous fishes).

ORDER i. ELASMOBRANCHII (sharks and skates),
Dev.-Rec.

2. PLEUROPTERYGII or CLADOSELACHII (sharks with
extremely primitive fins), Dev.-Carb.

3. PLEURACANTHODII or ICTHYOTOMI (early sharks
with continuous median fins and paired fins of Cemtodus
type), Dev.-Perm.

4. ACANTHODII (sharks with very strong spines —
ichthyodorulites — in the front of each fin), Sil.-Perm.

SUB-CLASS 3. Osteichthyes (bony fishes).
ORDER i. DIPNOI (lung fishes), Dev.-Rec.

2. ARTHRODIRA (armoured fishes, e.g., Coccosteus), Dev.-
Carb.

3. TELEOSTOMI (ganoids and teleosts), Dev.-Rec.

THE VERTEBRATA 237

CLASS: AMPHIBIA.

ORDER i. STEGOCEPHALIA (with armoured skulls),
Carb.-Perm.

2. URODELA (newts), Lower Cret.-Rec.

3. GYMNOPHIONA or CCECILIA (snake-like), Recent
only.

4. ANURA (frogs), Eoc.-Rec,

CLASS: REPTILIA.

ORDER i. ANOMODONTIA or THEROMORPHA (related to
Stegocephalia on the one hand and Mammalia on the
other). The dominant land-animals of the Permian
period, especially in the Southern Hemisphere. Perm-
Trias.

2. CHELONIA (tortoises and turtles), Trias.-Rec.

3. ICHTHYOPTERYGIA (ichthyosaurs), Trias. -Cret.

4. SAUROPTERYGIA (plesiosaurs, swimming reptiles,
less fish-like in aspect than the ichthyosaurs), Trias.-Rec.

5. RHYNCHOCEPHALIA (lizard - like, with biconcave
vertebrae, two rows of teeth in upper jaw, between which
the single row of lower jaw bites, all teeth fused to the
bones : the sole modern survivor, Hatttria, is confined
to a few of the smaller islands of New Zealand),
Perm.-Rec.

6. SQUAMATA (lizards and snakes), Upper Jur.-Rec.

7. DINOSAURIA (the dominant land-animals of the
Mesozoic after the anomodonts had decayed; a very
varied series, showing considerable adaptative radiation ;
some, as Diplodocus and Iguanodon, attained a great size ;
related to crocodiles and birds), Trias.-Cret.

8. CROCODILIA (crocodiles ; the earlier forms were
marine), Trias.-Rec.

9. AVES (birds), Upper Jur.-Rec.

10. ORNITHOSAURIA (pterodactyls), Jur.-Cret.

238 PALEONTOLOGY

CLASS: MAMMALIA.

SUB-CLASS : Prototheria (egg-laying mammals).

ORDER i. MULTITUBERCULATA (known only by lower
jaws and teeth), Trias. -Eocene.

2. MONOTRE.MATA (the egg-laying mammals of
Australia), Plio.-Rec.

SUB-CLASS : Metatheria (pouched mammals).

ORDER. MARSUPIALIA (now confined to Australasia and
South America), Jur.-Rec.

SUB-CLASS. Eutheria (placental mammals).

SUPER-ORDER i. UNGUICULATA (clawed mammals,
including the most primitive Eutheria, Insectivores and
Rodents, the Carnivora, and the two highly specialized
orders, the Edentates and the Bats), Eoc.-Rec.

2. UNGULATA (hoofed mammals, divided by Osborn
according to their geographical origin — Artiodactyla,
Perissodactyla, and several extinct orders being Hoi-
arctic,* Proboscidea, Hyracoidea, Sirenia [sea-cows], and
some extinct orders African, while a number of extinct
orders are South American).

3. PRIMATES (lemurs, monkeys, and man), Eoc.-Rec.

4. CETACEA (whales and dolphins, derived from primi-
tive carnivores), Eoc.-Rec.

Short Bibliography.

AGASSIZ, L. — Recherches sur les Poissons Fossiles,
1833-44.

DEAN, B. — Fishes, Living and Fossil. New York, 1895.

FLOWER, W. H., AND LYDEKKER, R. — Mammals,
Living and Extinct. London, 1891.

* The Holarctic Region includes the Old World north of the
Himalayas and the great deserts, and North America.

THE VERTEBRATA 239

GOODRICH, E. S. — Cyclostomes and Fishes, in Lan-
kester's Treatise on Zoology, pt. ix., fasc. i. London, 1909.

OSBORN, H. F. — The Age of Mammals. New
York, 1910.

WOODWARD, A. S. — Outlines of Vertebrate Palaeon-
tology. Cambridge, 1898.

PAL/EONTOGRAPHICAL SOCIETY'S MONOGRAPHS.

LANKESTER AND TRAQUAIR. — Old Red Sandstone
Fishes.

TRAQUAIR, R. H. — Carboniferous Ganoids.

WOODWARD, A. S. — (i)-Cretaceous Fishes. (2) Wealden
and Purbeck Fishes (in progress).

OWEN, R. — (i) Several Monographs on Mesozoic and
London Clay Reptilia. (2) Mesozoic Mammalia.
(3) Red Crag Cetacea.

ADAMS, A. LEITH. — Elephants.

DAWKINS AND SANFORD. — Pleistocene Mammalia,
vol. i.

REYNOLDS, S. H. — Pleistocene Mammalia, vol. ii.

BRITISH MUSEUM CATALOGUES.

WOODWARD, A. S. — Fossil Fishes, 3 vols.

LYDEKKER, R. — Fossil Amphibia and Reptilia, 4 vols. ;
Fossil Birds, i vol. ; Fossil Mammalia, 5 vols.

ANDREWS, C. W. — Tertiary Vertebrata of Fayum,
i vol.

VII
THE ECHINODERMATA

WHEN a piece of limestone is broken with a hammer it
usually shows a smooth or granular fracture, but here
and there on the broken surface may often be seen
distinct cleavage-surfaces of calcite : these may be few
or many, and in particular cases may attain a large size.
They are the fractured surfaces of fossil echinoderms.
Whereas the shells of most invertebrates show on a
broken surface a fibrous or laminated texture, due to the
arrangement of an indefinite number of minute crystals of
calcite, it is one of the essential features of echinoderms
(and of no other group except the calcareous sponges) that
the skeleton is composed of a series of units that may
attain a large size, but of which every one is, mineralogi-
cally, a crystal of calcite — not in external shape (which is
determined by organic secretion), but in molecular con-
stitution. These units have different names, according to
shape : the commonest are broad, flat structures, called
plates. In these the normal to the surface corresponds to
the vertical crystallographic axis. When the shape is
long and cylindrical, as in the radioles (articulated spines)
of sea-urchins, it is the long axis which is the vertical
crystal-axis.

240

THE ECHINODERMATA 241

Solid crystalline calcite would be much too heavy a
material for an organic skeleton. Accordingly, in the
living state the plates are lightened by rounded cavities,
arranged so closely as to make the whole substance
spongy, and in a very definite organic pattern. This
network of calcite is called the stereom. When the
skeleton of a recent echinoderm — say, the radiole of a sea-
urchin — is broken, the calcite cleavage is interrupted by
these abundant cavities, and is easily overlooked, though
careful observation will always show it. But the first
process in fossilization is always the filling-in of these
cavities with calcite in crystalline continuity with that
around them. The pattern of the cavities may or may
not be obliterated in the process ; when not, it forms
a further means of recognition (besides the crystalline
character and cleavage) of echinoderm skeletons in
microscopic sections of rocks. Calcite in crystalline con-
tinuity may also be deposited around the plates (or other
units), if there is opportunity, concealing the organic
form and replacing it, as far as circumstances permit, by
the geometrical form of a crystal. Thus, on breaking
open one of the sea-urchins so common in the Upper
Chalk, one occasionally finds that each plate projects
into the central cavity as a rhombohedron ; externally
the contact of the chalk has prevented such growth, and
if the interior has been filled with chalk (the more usual
case) the plates will preserve their original surface in-
ternally also.

There is another very general, but not universal,
feature characteristic of echinoderms — a five-rayed or

16

242 PALEONTOLOGY

pentamerous symmetry. Such a symmetry is common
among flowers, but is quite unknown among animals
except in the echinoderms. There is one other animal
phylum with a radial symmetry — the Coelenterata — and
in Cuvier's classification, now a century old, the two
were united as Radiata. But not only is the symmetry
of the Ccelenterata a four-, six-, or eight-rayed symmetry,
but it is in general a more perfect symmetry than that of
echinoderms. The former include some forms, such as
the jelly-fish, which are absolutely radial throughout
their bodies ; the latter include none such, though some
attain almost perfect radiality in their skeleton only. In
some, again, radial symmetry has not yet been attained ;
in others it has been modified into a practically bilateral
symmetry.

The skeleton of echinoderms is not truly external like
that of Brachiopods or most Mollusca ; there are soft
living tissues outside it, but these are so thin in most
cases that it may be roughly spoken of as external. Still,
it is important to remember these external tissues in
order to form an idea of the life of echinoderms.

The Echinodermata fall into two main divisions —
Pelmatozoa or fixed forms, and Eleutherozoa or freely-
moving forms. To be precise, the former are fixed at
some time in their life (and usually throughout adult
life), the latter never. The former were abundant in the
Palaeozoic era, rare afterwards ; with the latter the
reverse is the case. They are further divided into
classes, thus :

THE ECHINODERMATA 243

PELMATOZOA.

Cystidea. — Extinct, Cambrian to Permian, but mainly
Lower Palaeozoic.

Blastoidea. — Extinct, Silurian to Permian, but
mainly Upper Palaeozoic.

Edrioasteroidea. — Extinct, Ordovician to Carbon-
iferous.

Crinoidea. — " Sea-lilies," Ordovician to Recent.

ELEUTHEROZOA.

Echinoidea. — " Sea-urchins," Ordovician to Recent.
Stelleroidea.— " Star-fish," Cambrian to Recent.
Holothuroidea. — " Sea-cucumbers," Cambrian to
Recent.

All echinoderms are marine.

It will be convenient to describe first some examples
of crinoids.

i. Cupressocrinus gracilis (Fig. 67) is one of many
crinoids found in the Middle Devonian " crinoid-bed " of
Gerolstein in the Eifel (Rhenish Prussia). It consists of
a root, stem, and* crown. By the first (not often preserved)
it was fixed to the sea-bottom, but this "root" had no
absorbent function like the root of a plant. The stem
lifted it up high above the mud : as preserved fossil
it consists of a large number of squarish discs piled up
into a column : these are the stem-ossicles or columnals.
Through each of them runs a central vertical tube (axial
canal) and four smaller peripheral canals. Adjaceyt
columnals were united, in life, by organic tissues, and the
surfaces of contact were roughened to give greater grip
or " key " : on the articular surface of an isolated ossicle

244 FAL/EONTOLOGY

this gives, with the canals, a pattern which is fairly
characteristic of the genus Cupressocvinus. During life
the slight amount of flexibility of the organic tissue of
each joint gave a suitable degree of flexibility to the
whole stem.

The crown contained the essential vital parts of the
animal. It is divided into the theca or calyx, which
is the " body " of the animal and directly articulated to
the stem, and five arms which, when the animal was
feeding were spread out widely, but in the fossil are
usually found tightly closed up. The part of the calyx
which is visible when the arms are closed is called the
dorsal cup; it is the ventral surface which is hidden.
The dorsal cup consists of two circlets of five plates each
and a single pentagonal plate next to the stem. This
last is regarded (by analogy with other crinoids) as a
fused third circlet, the five plates of which if separate
would be called infrabasals ; the five plates in the circlet
next to it are called basals ; those in the upper circlet,
radials. The straight lines along which adjacent plates
meet are termed sutures. Radials and basals alternate in
position, and if the infrabasals were separate they would
alternate with the basals. Because of the presence of
(fused) infrabasals as well as basals, this crinoid is said
to be dicyclic. In many crinoids the top columnal
directly joins the basals : they are monocyclic.

The five radii which can be drawn from the centre of
the calyx through the middle of each radial plate are
called perradii, and organs such as the arms which lie
symmetrically upon them are said to be perradial in posi-

THE ECHINODERMATA

245

tion. The radii half-way between these are called
inUrradti, and organs which lie upon them are said to be
intervadial (e.g., the basal plates). These alternating
positions are of fundamental importance in all echino-
derms, except the lowest forms in which five-rayed

FIG. 67. — CUPRESSOCRINUS.

a, C. gracilis Goldfuss ; b, c, C. abbreviates Goldfuss ; both Middle
Devonian, Eifel. a, Side view of complete crown, with arms closed,
and part of stem ; b, oral view ; c, basal view of a cup without arms or
stem. (All natural size.) (Original.) A, Anus; A.C., axial canal;
Ax, axial canal of stem ; B, basal plate ; Br. i, 2, first and second
brachial plates ; IB, infra-basal plate ; M, mouth ; O, oral plate ;
R. , radial plate.

symmetry is not fully developed. It is a universal rule
that the mouth and anus, unless they are central, are
respectively perradial and interradial in position. With
the mouth, however, the central position is the rule, with
the anus it is the exception.

246 PALEONTOLOGY

In the case of Cnpressocrinus the mouth is central and
appears as a large circular opening between five large
plates, interradial in position. Probably the actual
mouth was smaller, and other smaller plates surrounded
it. The five large plates are called orals or deltoids : one
of them differs from the rest in being deeply notched by
the anus. This is the only disturbance of radial
symmetry in the crown of Cupressocrinus, and it marks out
the odd radial as posterior in position ; the two adjacent
to it as postero-lateral and the other two as antero-
lateral. The margins of the orals that are in contact
with their neighbours bear semi-oval notches, which com-
bine into five pear-shaped perforations corresponding to
the middle line of each arm. Along these were continued
the five food-grooves that occupy the middle ventral line
of the arms. As the edges of the five orals meet around
the mouth the food-grooves must have finally entered the
mouth by tunnels, and the pear-shaped openings them-
selves may have been converted into tunnels by small
overlapping covering-plates (ambulacrals).

Each arm consists externally (dorsally) of a single row
of large bmchial plates, tapering to the free end. In-
ternally (ventrally) they have a wide and deep groove,
with a row of pinnules (like miniature arms) along each
side, and a series of small plates (ambulacrals) arching
over it and converting it into a tunnel. This groove, in
life, was ciliated and constituted the food-groove or
ambulacrum, along which microscopic food was wafted
towards the mouth. The pinnules were similarly grooved
and ciliated, and served as tributaries. Thus the method

THE ECHINODERMATA

247

of feeding (microphagous) is fundamentally the same as in
brachiopods and lamellibranchs, but the structures
adapted to the same ends are fundamentally different.

Not to be confused with the food-grooves are longi-
tudinal cavities, elliptical in section, in the substance of
the brachial and radial plates : these are the axial canals
and contained a nerve-cord.

iBr.

Br.

Ax,

-A.F.

a.

Br.

FIG. 68. — AMPHORACRINUS ATLAS (M'Cov), VISEAN (CARBON-
IFEROUS LIMESTONE), YORKSHIRE.

a, Side view ; b, base. (Natural size.) (Original.) A, Anal plate inter-
calated between radials ; A.F., facet to which arm is articulated;
Ax.t axillary plate (primaxil) ; Br. , brachial plate (primibrach) ;
i Br., interbrachial plate (inter-primibrach) ; R, radial plate.

2. Amphoracrinus atlas (Fig. 68) is a not uncommon
fossil in the Carboniferous Limestone knolls of Clitheroe
and elsewhere in the North of England. Usually the
calyx is found without arms or stem. The columnals are
abundant, though not with certainty distinguishable from
those of allied forms. Their articular surfaces show fine
radiating lines and a pentagonal axial canal.

There are no infrabasals (monocyclic) and the basals
are reduced to three by crowding, leading to fusion. The

248 PAL/EONTOLOGY

radials are five in number, but when we have identified
the basals and these we still have a great many plates
in the dorsal cup unaccounted for. Nevertheless, we
cannot have mistaken the radials, because there are
plates directly above them which lead up to the arms,
and that could never be the case with basals. These
plates above correspond with the lower brachials of
Cupressocrinus, but instead of forming part of free arms
they are incorporated in the dorsal cup : they are called
fixed brachials. But holding them fixed are other plates,
interradial in position, arranged in bifurcating series :
these are called interbrachials. The largest of them dis-
turbs the five-rayed symmetry by pushing itself between
the radials and meeting two of the basals : it is called
the anal, not because it carries the anus, but because it
forms the base of the interray, high up in which lies the
anus.

The ventral surface, instead of being flat and of few
plates, as in Cupressocrinus, is a lofty dome of many
^ plates, solidly articulated together. Crinoids with this
:eature are confined to the Palaeozoic, and are known as
Camerata. This arrangement closes in the mouth com-
pletely and shuts it off from all possibility of the excre-
ment from the anus fouling the food. As an additional
precaution, the anus is lifted up from the level of the
tegmen by a short anal tube.

The arms branch into two immediately they become
free from the dorsal cup, and again branch repeatedly.

3. Marsupites testudinarius (Fig. 69) is a crinoid
characteristic of a definite zone high up in the White

•

THE ECHINODERMATA 249

Chalk. It is one of the few crinoids that were not fixed
in the adult stage. By analogy with the modern free
crinoid Antedon we may infer that it was stalked in a
larval stage, but of this there is no actual evidence. The
top columnal, however, is apparently retained as a large
pentagonal plate in the centre of the dorsal cup. Around
and above this are five large pentagonal infrabasals, then

A.F.

A.F.

A.F:

C.D.
FIG. 69.— MARSUPITES TESTUDINARIUS (SCHLOTHEIM), CAMPANIAN

(UPPER CHALK).
(Cup without arms, natural size.) (Original.)

A.F., Arm-facets; B, basal plates; C.D., centro-dorsal plate ; IB, infra-
basals ; R, radials.

five large hexagonal basals, and five large pentagonal
radials, with facets for the arms. So far as the dorsal
cup goes there is perfect five-rayed symmetry, and the
dicyclic condition is shown very simply and clearly.

The arms branch repeatedly. The ventral surface is
never preserved, but we may assume that, as in other
post-Palaeozoic crinoids, it consisted of many small
plates forming a flexible surface, on which lay open food-

250 PALEONTOLOGY

grooves, converging to an exposed mouth. An inter-
radial anus would be the only disturbance of symmetry.

Isolated plates of Marsupites can easily be recognized
by their curious ornament of ridges arranged in sets at
right angles to the sutures. An almost identical pattern
is found, however, in the Silurian Crotalocrinus.

There is much difference of opinion as to the classifi-
cation of crinoids. Dr. Bather, of the British Museum,
regards the distinction between monocyclic and dicyclic
forms as fundamental, whilst most other authors treat it
as of merely family value. However sound this distinc-
tion may be on morphological grounds, it has the
practical difficulty that many dicyclic crinoids are
" pseudo-monocyclic," so that the ordinary student does
not recognize their dicyclic character.

In both monocyclic and dicyclic forms certain definite
grades of structural complexity are found, which are
taken as a basis for division into orders: (i) Inadunata,
in which there is no incorporation of fixed brachials in
r the dorsal cup ; (2) Camerata, with the tegmen in the
form of a rigid vault covering the mouth and food-
grooves (Palaeozoic only); (3) Flexibilia or Articulata,
in which the tegmen is composed of small plates loosely
articulated, and the mouth and food-grooves exposed
(almost exclusively Mesozoic and later). But authorities
differ considerably as to the limits of these orders.

Among the simplest Inadunata (Larviformia of Wachs-
muth and Springer) is the dicyclic Cupressocrinus, already
described. Pisocrinus (Fig. .71, a, Sil.) is a monocyclic of
equally simple structure, but much smaller sized cup,
though the very slender arms attain as great a length as
those of Cupressocvinus ; the cup-plates have their sym-
metry disturbed by the subdivision of the right posterior
radial obliquely into two plates (r.p. and R') — a common

THE ECHTNODERMATA

251

feature in many inadunates. Herpetocrinus (Sil.) is found
with the stem spirally coiled, concealing the crown, but
it could straighten itself when such protection was not
necessary.

Platycrinus (Fig. 71, b, Carb.) is a monocyclic, transi-

FlG. 70,-^-CRINOID COLUMNALS.

#. b, e, f, g, Views of articular surfaces ; c, vertical section of part of stem ;
d, internal cast in place in rock-matrix ; e', side view. The thick black
in a-d represents cavities. (All about natural size.) (After Goldfuss.)
<z, Actinocrinus granulatus Goldfuss, Devonian, b, Cyathocrinus
rugosus Milkr, Devonian, c, d, Cyathocrinus pinnatus Goldfuss,
Devonian, e, e', Encrinus moniliformis Miller, Triassic. ftlsocrinus
basaltiformis (Miller), Jurassic, g, Bourgueticrinus ellipticus (Miller),
Upper Cretaceous (Upper Chalk). Dotted line = long axis of articular
surface below.

tional from Inadunata to Camerata. The basals are re-
duced to three by fusion ; the symmetry is almost perfect,
there being only a slight difference between the anal
interray and the others ; the arms branch repeatedly, and
have a large crescentic articulation on the radials. The

252

PAL/EONTOLOGY

O-

^[^^fer-i

V UAytV/ - 1«
a O O

rp r.a. I. a.

FIG. 71.— CRINOIDEA.

IB, Infrabasals; B, basals ; R, radials ; Ax, axillary, a, Pisocrinus,
dissected. (After Either.) a, Anterior ; />, posterior ; /, left ; r, right ;
R , radianal ; xt anal, b, Platycrinus Icevis Miller, Lower Carbon-
ifero^. Cup only. (After de Koninck.) A.F., Arm -facet, c, Crota-
locrmts ruifosus (Miller), Silurian (Wenlock Limestone). Cup only.

THE ECHINODERMATA 253

columnals have elliptical surfaces, but the long axis of
the ellipse on the lower surface is at an angle to that on
the upper surface. Thus the axis of bending of succes-
sive joints gradually changes, enabling the stem as a
whole to move in any direction in spite of the shape of
the columnals. This arrangement gives the stem a
curious twisted appearance. Good specimens are found
in the Carboniferous Limestone of England and Ireland.

Marsipocrinus (Sil.) differs in the very rapid branching
of each arm into four, the parts below the separation
being united by single interbrachial plates, so that much
more of the brachials are incorporated in the cup.

Crotalocrinus (Fig. 71, c, Sil.) is a dicyclic inadunate in
which the arms branch with great suddenness, and all the
branches of one ray are united into a sort of membrane,
which was spread out horizontally when the crinoid was
feeding at ease and rolled up spirally on an alarm. The
stem is as broad as the crown, is composed of very wide
and thin columnals, and has a very rough surface.

Cyathocrinus (Fig. 71, g, Ord.-Carb.) and Gissocrinus
(Sil.) are dicyclic inadunates with very simple cup, one
anal only intercalated, and much-branched arms. In the
former genus the arms branch in a rather lax and irregular
manner, and the anal tube is often very long; in the

FIG 71. — CRINOIDEA (continued}.

(After Murchison; restored.) d, Taxocrinus tuberculatus (Miller),
Silurian (Wenlock Limestone). (After Murchison.) Br. i, First
brachial ; i Br., interbrachial. e, Eucalyptocrinus decorus (Phillips),
Silurian (Wenlock Limestone). (After Murchison ; restored.) IBr.i..,
First primibrach ; I Ax., primaxil; //Br.i., first secundibrach ;
ilBr., inter-primibrach ; A7, niche; V.S., visceral sac. /, Woodo-
crinus expansus de Koninck (Carboniferous Limestone), Yorkshire.
(After Roberts.) Pin. , pinnules, g, Cyathocrinus acinotubus Angelin,
Silurian (Wenlock Limestone). (After Bather.) h, Apiocrinus par
kinsoni Schlotheim, Bathonian (Bradford Clay). Cup and part of
stem, and of arms. (After Goldfuss.) Prox., proximal columnal:
I Br. , primibrachs ; 7/^r., secundibrachs. i, Rncrinus monihformis
Miller, Ladinic (Muschelkalk). Brachials shown on lower part of one
arm only. (After Goldfuss.) /', base ; Col., proximal columnals ;
// Br. i. , first secundibrach. (After Goldfuss. ) (All X J, except a.)

254 PALEONTOLOGY

latter the arm-branches are closely packed together, the
anal tube is compressed and its plates are wider
than high.

Petalocrinus (Sil., marking a definite zone) has the
branches of each arm united into a fan-like body.

Woodocrinns (Fig. yi,/), of which beautiful examples
are found in the Carboniferous Limestone of the North
of England, is a dicyclic inadunate with four-branched
short arms and a relatively short stem tapering to a point.

Among Camerata, Amphoracrinus has been described.
Actinocrinus is similar, but with the superficial area of the
tegmen not larger than that of the cup, and a longer and
central anal tube. A very remarkable genus is Euca-
lyptocrinus (Fig. 71, e, Sil.-Dev.), with the base deeply
concave, the visible part of the cup being mainly com-
posed of brachials and interbrachials : the arms bifurcate
within the cup, and from the ten sets of interbrachials
there grow up vertical pillars which arch out and meet
at their upper ends, forming ten niches in which the
arms lie when not spread out. There are a few dicyclic
camerates, of which Rhipidocvinus (Dev.) and Rhodocrinus
(Carb.) may be mentioned.

The few Palaeozoic Flexibilia are dicyclic and impin-
nate (i.e., arms have no pinnules) : Taxocvinus (Fig. 71, dt
Sil.-Carb.) is the commonest. The bes{-known Triassic
crinoid is Encvimis (Figs. 70, e ; 71, j), which seems to be
related to Carboniferous inadunates such as Woodocrinus.
It is dicyclic, but the infrabasals are minute and difficult
to find.

Isocrinus (Fig. 70, /, Trias.-Rec.) and Pentacrinus (Jur.)
not only have minute infrabasals, but the basals are very
small, and the radials intervene between them and come
into contact with the stem, so that the whole cup, lying
between a large stem and large, much-branched and
pinnulate arms, is reduced to insignificance. The stem

THE ECHINODERMATA 255

is five-sided, and the columnals show a very striking
pattern. Apiocrinus (Fig. 71, h, Jur.) has a cylindrical
stem, which expands at its upper end, so that its outline
passes imperceptibly into that of the crown ; the infra-
basals are not recognizable (pseudo-monocyclic or crypto-
dicyclic), the arms branch into two very quickly, and
their lower part is incorporated in the cup. Beautiful
specimens have been found, and fragments are common,
in the Bradford Clay of Bradford -on- Avon. In Bouv-
gueticriniis (Fig. 70, g, Cret.) most of the columnals have
the same peculiarity as in Platycrinus (ante, p. 251), but are
longer and more barrel-shaped ; towards the crown they
are round and the proximal one is as wide as the cup,
which shows five basals and five radials : the arms are
rarely preserved. Antedon or Comatula (Jur.-Rec.) is
fixed by a stem in its earliest youth, but afterwards
becomes free-swimming. Another free-swimming form is
Saccocoma, in which the radials form nearly the whole
cup ; the arms are bifurcated and the ten branches
often spirally coiled, the brachials bearing wing-like
expansions. This is abundant in the Upper Jurassic
lithographic stone of Solnhofen, Bavaria, and pyritized
plates are found at the same horizon in the English
Kimmeridge Clay, recognizable by the coarse meshwork
of stereom.

The Cystidea are a somewhat heterogeneous group,
consisting of the most primitive forms, in which gradual
development of five-rayed symmetry can to some extent
be traced.

Echinosphaera aurantium (Fig. 72) is very abundant
in the Cystid Limestone of the Ordovician of Sweden,
erratics of which, carried by ice, are common on the
plains of North Germany. It is an almost spherical

256 PAL/EONTOLOGY

body, a slight projection at one point forming a rudi-
mentary stem, while almost opposite it is the slightly-
raised rim of a circular opening, presumably the mouth.
About one-third the distance from the mouth to the stem
is a third projection — a low pyramid of five triangular
plates, the anal pyramid. Between this and the mouth,
but nearer the latter, is a very small round opening,
interpreted as a kydropore, that is an opening by which
water is taken into the system of water- vessels.

FlG. 72. — ECHINOSPH^RA AURANTIUM, GVLLENHALL, ORDOVICIAN,

SWEDEN.
(Natural size. Original.) (Only a few of the plates are indicated. )

A, Anal pyramid ; H, hydropore ; M, mouth.

The spherical theca is made up of a very large number
of polygonal (mostly hexagonal) thecal plates, which when
closely examined show a pattern somewhat resembling
that on the infrabasals of Marsupites, though due, not as
in that to external ornament, but to the internal structure
of the plates. The stereom of each plate is thrown into
folds perpendicular to the adjacent suture, so that the
whole surface appears divided up into a series of rhombs,
each belonging half to one plate and half to another,
the folds in each rhomb forming a parallel series. The

THE ECHINOD£RMATA 257

cystids showing this structure form an order Rhom-
bifera.

In specimens much better preserved than usual small

M

ABr

AA

FIG. 73. — CYSTOIDEA.

a, Dendrocystis scotica Bather, Ordovician, Girvan. ( X |.) Details of plates
omitted from brachiole. (After Bather.) b, Lepadocrinus quadri-
fasciatus (Pearce), Silurian (Wenlock Limestone). Slightly enlarged ;
details of plates on arms omitted. (After Forbes.) c, Glyptosphcera
leuchtmbergi Volborth, Ordovician, Pulkowa (Russia). ( X £. ) Mouth
in centre, with five radiating grooves, irregularly branched, each
branch ending in articulation of a brachiole (represented by black
spots : where plate-sutures are shown, each of these is seen to be in the
centre of a plate). Anus below, black (plates of pyramid missing).
Thecal plates only shown around anus. (After Jaekel.) A, Ana)
pyramid; A A, anti-anal process; ABr, anti-brachial process; Br.,
brachiole; M, mouth ; P.R., pectinirhomb.

arms have been found arising from the mouth, composed
of grooved brachial and smaller ambulacral plates.

17

258 PAL/EONTOLOGY

These are usually three in number, but may be two or
four. Thus there is nothing in Echinosphara showing
pentamerous symmetry, unless it be the anal pyramid.
Certain other cystids, however, give a hint of how the
pentamerous condition arose : they have also three food-
grooves, but while the anterior one (opposite the anus)
remains unbranched, the other two branches soon divide
into two, giving five grooves with an imperfect radial
symmetry (Fig. 73, c).

The Cystidea are divided into four orders :

1. Amphoridea, with no trace of radial symmetry,
varying from simple sacs like Aristocystis (Ord. of
Bohemia) to forms with well-developed stem, flattened
theca and more or less bilateral symmetry. Dendvocystis
(Fig. 73, a, Ord.) has one long arm, with the mouth at
its base, the anus at the opposite end near the stem, both
on the same side, while on the other side are outgrowths
(A Br and A A) which served to balance them. There
are still more specialized bilateral forms like Placocystis
(Sil.) with a curious mimicry of Branchiopod Crustaceans,
which suggests that they were actively moving animals.

2. Rhombifera, showing the rhomb - foldings of
stereom described above. Echinosphara is about the
simplest of these. Others like Macrocystella (Tremadoc)
show stem, circlets of thecal plates, and arms almost like
crinoids. Others like Lepadocrinus (Fig. 73, b, Sil.) show
a remarkable specialization of a few (usually three) of
the rhombs into large pectinirhombs with all the appear-
ance of respiratory organs (gills), while the food-grooves
are on the crests of ridges which might be described as
arms sessile on the theca.

3. Aporita, a small group with fair pentamerous
symmetry and no special stereom-structures, e.g. Crypto-
crinus (Ord.)

THE ECHINODERMATA 259

4. Diploporita, in which the stereom-cavities are
normal to the surface and grouped in pairs, giving rise to
an appearance which must by no means be confused
with the very different " pore-pairs " of Echinoidea.
There is fair pentamerous symmetry, and the food-
grooves radiate on the surface of the theca. SpharoniUs
(Ord.) and Glyptosphcera (Fig. 73, c, Ord.) are spheroidal
contemporaries of Echinospha'ra.

The Blastoidea may be illustrated by Pentnmites
pyyifovmis (Fig. 74), of the Lower Carboniferous of North

FIG. 74.— PENTREMITES PYRIFORMIS, SAY, MISSISSIPPIAN
(CHESTER GROUP), ALABAMA, U.S.A.

a, Side view. (Natural size.) t>, Oral region. (X2.) (Original.) B, Basals ;
R, radials ; L, lancet- plates ; O, orals or deltoids. In &, the mouth is
in the centre, the five spiracles around it, and beyond and alternating
with these the five ambulacra.

America. The shape suggests a flower-bud just opening.
There is no true stem, but the base is drawn out into a
stem-like stump. It would consist of five lozenge-shaped
basals, but two pairs are fused, so that only three basals
are counted. The five radials are much larger and would
be hexagonal but for a very deep notch which almost
cuts each in two. Five small orals or deltoids alternate
with the radials above. Between these, and in the
notches of the radials, lie five lancet-plates, each with a
central food-groove into which run numerous oblique

260 PALEONTOLOGY

grooves from right and left alternately. Between the
margins of the lancet-plates and the radials are side-plates,
which in exceptionally preserved specimens bear each a
pinnule. From their position in relation to other
structures it is obvious that a lancet-plate with its side-
plates (sometimes termed an ambulacrum) answers to the
arm of a crinoid— an arm which is sessile upon and forms
part of the calyx. The side-plates bear pores, which
open into internal gill-like organs, the hydrospires. The
mouth is in the centre of the upper surface : around
it are five openings, interradial in position, divided into
two internally, the spiracles, which also communicate
with the hydrospires. Each spiracle is roofed by the
halves of two lancet-plates and their side-plates. The
anus is probably confluent with one of the spiracles.

In another common American species, P. godoni, the
base is flattened at the level of the lower end of the
lancet-plates. In Orbitremites, found in the Carboniferous
Limestone of England, but not common, the calyx is
ellipsoidal, and the ambulacra very narrow and extending
very far down towards the base.

The Eleutherozoa are always freely-moving benthic
animals, sometimes carnivorous, sometimes mud-eaters,
but never microphagous as are all Pelmatozoa (even the
few nectic forms)

Hemicidaris intermedia (Fig. 75) is a sea-urchin of
which many very perfect specimens have been found in
the Coral Rag of Calne in Wiltshire. The general
characters and symmetry of the skeleton suggest at once

THE ECHINODERMATA 261

an affinity to crinoids, but there are no stalk and no arms.
The skeleton (or test) has somewhat the appearance of a
globe with poles flattened, but one (the oral pole) so much
more so than the other (aboral) that the flattened surface
is but little way below the equator (or ambitus). A full-
grown specimen is about 36 mm. in diameter at the ambitus
and 25 mm. high. Sometimes the fossil is found with
the movable spines (or radicles) which characterize the
sea-urchins still in their natural position ; but as a rule
they fell off before burial, and the test only shows the
tubercles with which they were articulated. Some of
these are large and prominent, and carried thick cylin-
drical radicles, up to 95 mm. in length. These large (or
primary) tubercles are arranged in five double rows,
between which are narrower areas each with a double
row of much smaller tubercles. The difference between
these alternate areas is most striking near the aboral
pole, and becomes much less so near the oral pole by the
increase in width of the narrower areas and in the size
of their tubercles. But there is a more important dis-
tinction that persists throughout: the narrower areas
carry a series of fine pores, always arranged in pairs, and
in each area there are two vertical columns of pore-pairs.
These pores, in life, transmitted finger-like muscular
tubes kept tense by the pressure of water in them from
an internal system of water-vessels (analogous to blood-
vessels) : these are the tube-feet (or podia), the organs
of locomotion of sea-urchins. In life they extend out
beyond the long radioles, and their sucker-like ends can
adhere firmly to any foreign body.

262

PALEONTOLOGY

The five areas containing pores are called the ambu-
lacral areas, or, briefly, the ambs ; the alternate areas are
the mterambulacral areas or interambs. The ambs are
perradial, and the interambs interradial. Together they
form the corona, which is the greater part, but not the
whole, of the test.

It is not always easy to see the sutures of the plates

FIG. 75. — HEMICIDARIS INTERMEDIA, FLEMING, ARGOVIAN
(CORAL RAG), CALNE (WILTS).

a, Aboral view. (x£.) All details of ambs and interambs omitted.
b, Apical disc. (Natural size. ) c, An interamb plate in profile. (X2.)
d, Two interamb plates with adjacent part of amb. ( X 2. ) e, Large
rndiole. (Xf.-) (After Wright.) A, Periproct ; Ac., acetabulum
(articular socket for mamelon) ; Amb., ambulacral area; B, boss;
G, genital plate; /.A., interambulacral area; M, mamelon; Mad.,
madreporite ; Oc. , ocular plate.

which build up the ten areas, especially on the external
surface. On the internal surface they are plainer. Each
area consists of two vertical columns of plates, the sutures
down the middle of an area being zigzag, while those
between adjacent areas are nearly straight. In the
interambs there is a plate to each large tubercle. In the
ambs there is, near the aboral pole, one pore-pair to each

THE ECHINODERMATA 263

plate, and each plate extends across half the width of the
amb. Such plates are called simple primaries. Lower
down compound plates appear, each with several pore-
pairs. The development of living forms shows that such
plates arise by the fusion of simple plates, and sometimes
the fused sutures can be traced in a fossil. When this
can be done in Hemicidaris, it is found that three plates
fuse together : the middle one carries most of the large
tubercle and may be the only one that occupies the full
width of its column (primary plate), the other two only
covering about two-thirds of that width (demi-plates).

Growth of the corona takes place (i) by intercalation
of new plates at the aboral end of the ambs and inter-
ambs, and (2) by increase of size (if necessary, with
alteration of shape) of existing plates. The early-formed
coronal plates thus get thrust farther from the apical
system as growth proceeds. Sometimes plates next the
peristome may undergo resorption. Thus, though in-
dividual plates grow by accretion, the corona as a whole
does not, but grows by a more complex process of intus-
susception. Consequently the way in which the adult
corona illustrates its own development is the reverse of
that seen in molluscs and brachiopods. In echinoids
the earliest-formed coronal plates, instead of illustrating
the earliest stage of growth, are the most highly de-
veloped ; while those which have been latest formed show
the simplest characters. Thus it is that in Hemicidaris
the ambulacral plates next to the apical disc are simple,
while those near the peristome are compound.

The ambs and interambs end orally around a large

264 PALEONTOLOGY

space, the outline of which has ten notches, which mark
the position of external gills. This space in fossils is
colloquially termed the mouth, but that is not strictly
correct, the real mouth being smaller, and surrounded by
a membrane (fieristome), the plates in which are loose and
on its decay fall away, leaving a large space. The mouth
was furnished in life with a complex series of biting
organs (jaws and teeth) forming the lantern of Aristotle.
This is rarely preserved, but its existence is shown by
the presence of the perignathic girdle, a series of internal
arches arising from the lowest plates of the corona, and
serving as an origin for the muscles inserted on the
lantern.

Towards the aboral pole the corona is surmounted by
a flat series of ten plates called the apical disc (or better,
apical system), in the centre of which is a space, collo-
quially called the anus, but more correctly the periproct
(for a similar reason to that of the peristome). Of
the ten plates, the five smaller are at the ends of the
ambs and are called the ocular plates ; the five larger
interradial plates are termed genital^ because they bear
the apertures of the genital ducts. One of the genital
plates, in addition to the genital pore, is perforated by a
great number of smaller pores, through which water
filters into the water-vessels: this plate, called the
madreporite, constitutes the only imperfection of five-
rayed symmetry in the test of Hemicidaris. The ocular
plates have been termed radials, and the genital, basals ;
but they are certainly not strictly homologous with the
plates so named in crinoids, which do not encircle

THE ECHINODERMATA 265

the anus: they seem more comparable with the plates
of the anal pyramid in cystids.

The radioles of Hemicidaris articulate with the tubercles
by a ball-and-socket joint. They are moved by muscles
in the skin, which lies outside the test. The tubercles are
distinguished according to size into primary, secondary,
and miliary. The primary tubercles consist of a boss
and a mamelon (Fig. 75, c) : in Hemicidaris the inner

FlG. 76.— CONULDS ALBOGALERDS, LESKE, EMSCHERIAN

(UPPER CHALK).

a, Side view. (X|.) b, Oral view. (Xj.) c, One interamb plate and
adjacent amb plates. (X2.) d, Apical disc. (X2.) (After Wright.)
Sutures of amb plates omitted in a; of all plates in b; madreporite
dotted in d.

part of the boss shows a ring of radial crtnulations, and
the centre of the mamelon has . a perforation ; so the
tubercles are described as cnnulate and perforate.

ConulllS albogalerus (Fig. 76) is a common sea-
urchin of the White Chalk. The first obvious differences
from Hemicidaris are its more conical shape and much
smoother surface. Next we notice that the symmetry
is not strictly radial, but is bilateral, the base being a
pentagon of which one side is highly convex, while the

266 PALEONTOLOGY

others are flat : in the middle of this side is a large
opening, the periproct. The mouth, however, is still in
the centre of the oral surface, and the very small apical
disc is equal central at the aboral apex.

The smoothness of the surface enables the sutures of
the plates to be traced easily. The interambs are broad
and composed of large, nearly oblong plates, with from
one to twelve or more larger flat tubercles and numerous,
much smaller tubercles (granules). The ambs are narrow ;
their plates are much smaller and more irregular, includ-
ing simple primaries, simple demi-plates, and some com-
pound plates ; each has one or sometimes two larger flat
tubercles, and a number of granules.

The peristome is small ; the lantern and perignathic
girdle are in a vestigial condition. The apical disc is
not very different from that of Hemicidaris, except that
the madreporite is larger than the other genital plates
and extends into the space vacated by the periproct.

Micraster cor-anguinum (Fig. 77) is another sea-
urchin from the Upper Chalk. Viewed from above or
below its outline is heart-shaped, the anterior amb being
deeply depressed so as to notch the outline. The other
ambs are also sunk in a similar way for some distance
from the apical disc, but not far enough to affect the
outline. The posterior interamb is raised into an obtuse
ridge, the keel or carina, which runs almost horizontally
back from the apical disc to the hind margin which
drops perpendicularly. Sometimes the carina projects
slightly above this vertical end, forming a rostrum. The
anus lies near the top of this vertical face.

THE ECHINODERMATA

267

The mouth lies far forward (four-fifths of the whole
length) and faces forwards instead of downwards, as a
distinct lower lip (labrum} projects on its under side.

In the apical disc the two posterior ocular plates are in
contact, the genital plates between them being crowded
out ; the madreporite is large and extends into the centre
of the disc, taking a flask-like shape and in contact with

FlG. 77. — MlCRASTER COR-ANGUINUM, LESKE, EMSCHERIAN
(UPPER CHALK).

a, Aboral view. (xJ.) 3, Oral view. (Xi-) c, Side view. (Xj.)
d, Apical disc. (X~4.) (After Wright.) Some of the interamb sutures
are indicated in a and If. A, anus; Car, carina ; Fa, sub-anal fasciole ;
Lab, labrum ; M, mouth ; Mad, madreporite ; PI, plastron.

all the other plates except the left posterior ocular.
Otherwise the plates retain a fairly regular arrangement.
The ambs are composed, near the apical disc, of close-
set plates, very short and broad, with uniserial pore-
pairs. At about half-way towards the ambitus a sudden
change takes place : the plates become much larger and
squarer, and their pore-pairs become much less notice-

268 PAL/EONTOLOGY

able. The three anterior ambs can be traced directly
down to the mouth ; the two posterior extend to the
posterior end of the plastron, whence they can be traced
with some difficulty along either side of the plastron. It
is evident that the abundant tube-feet of the dorsal surface
cannot serve to convey food to the mouth, since they are
separated from it by a wide space with very scanty tube-
feet : probably they serve as gills. These aboral regions
of abundant pore-pairs are called the petals, from a
resemblance to an open five-petalled flower ; but Micrastey
is not typically " petaloid," only sub-petaloid, because the
two series of pores in each petal are nearly parallel,
instead of widening out and closing in again.

The interamb plates, having to occupy all the space
between the ambs, are for the most part very large.
Their regular double-row becomes rather confused on
the under side. The posterior interamb on the under
side constitutes the plastron.

All plates except those of the petals bear tubercles and
granules. On the upper surface the tubercles are small,
not very prominent, and scattered, with abundant small
granules between. On the under surface they are larger,
with more deeply excavated areola, and are much more
closely set ; the granules, though fewer in number, being
also coarser and more close-set. Running round the
lower corner of the vertical posterior end there is a
narrow ring which at first sight appears smooth, but
under magnification is seen to be covered with extremely
fine granules. This is a fascicle, and from its position is
called a sub-anal fascicle.

THE ECHINODERMATA 269

The micrasters of the White Chalk, of which M. cor-
anguinum is the latest, are famous from the work of Dr.
Rowe, who showed in 1899, from the study of large
numbers of specimens collected and arranged with care
according to their exact zones, that they went through
a gradual series of changes which did not bear a close
relation to the accepted distinctions of species. These
changes, in fact, run on parallel lines in separate stocks,
so that they may serve as zonal indices. The chief
changes may be summarized thus :

(1) In general form they tend with the lapse of time
to become slightly broader and very distinctly higher,
while the ambital outline tends to change from wedge-
shape to ovate.

(2) The carina appears and becomes more marked, the
highest point which was at first at the apical disc shifting
backwards ; finally a rostrum appears.

(3) The anterior groove becomes deeper.

(4) The mouth shifts forward and the labrum makes
its appearance; at first smooth at the tip, it becomes
more and more granulated; the labral plate changes
from triangular to oblong, and its tubercles from irregular
to linear in arrangement.

(5) The amb-petals become longer and shallower.

(6) The sub-anal fasciole, from being weak, becomes
very strong, and loses a slight tendency towards an
8-shape which it had at first.

(7) The periplastronal area (the two posterior ambson
the under surface) from smooth passes to granulated and
finally to very coarsely granulated (mammillated).

270 PALAEONTOLOGY

(8) Perhaps the most easily observed change of all is
that in the interpoviferous areas (i.e., the middle strip of the
petals, between the pore-pairs). These pass through a
series of stages (a) " smooth " ; (b) " sutured," i.e. the
sutures of the plates are clearly marked; (c) " inflated,"
i.e. instead of being gently concave, they rise in a double
convexity ; (d) " subdivided," i.e. the middle line between
these convexities becomes a distinct groove ; and
(e) " divided," i.e. this groove becomes very deep and
narrow.

The main division in the classification of Echinoidea is
between those with radial and bilateral symmetry, or
Regularia and Irregularia. The great differences between
these two sub-classes may be tabulated thus :

REGULARIA. IRREGULARIA.

Radial symmetry, perfect ex- Bilateral symmetry, perfect ex-
cept for madreporite. cept for apical system.
Anus within apical system, Anus in posterior interamb.
which is symmetrical and Apical system usually asym-
large. metric ; small and compact,

or elongated.

Mouth central, with well-de- Mouth either central (when jaws
veloped jaws. may be well- developed or re-

duced), or in anterior amb
(when jaws are lost).

Corona usually bearing large, Corona hardly ever with large,
prominent tubercles (as well conspicuous tubercles. Radi
as smaller ones) with long oles small,
stout radicles.

The Palaeozoic echinoids are distinguished from all
later forms (except one Triassic and one Barremian
genus) by the inconstancy in the number of coronal
columns of plates, which may be more or less than
twenty, but never that exact number, which is the in-
variable number among later forms (with the one excep-

THE ECHINODERMATA 271

tion named). This has been made the basis of a
classification into " Palechinoidea " and " Euechinoidea,"
but as a morphological classification this leaves Tetvaci-
daris in an anomalous position; as a genetic classification
it is unsatisfactory, as it unites a number of divergent
Palaeozoic stocks, of which one was apparently more
closely related to Mesozoic regular echinoids than the
others. The following classification is based on that of
Martin Duncan, with modifications taken from the
classifications of Professor J. W. Gregory and Dr. R. T.
Jackson, neither of which is fully adopted.

CLASS: ECHINOIDEA.
SUB-CLASS : Regularia (Endocyclica).

Order i. BOTHRIOCIDAROIDA. With a single column
of interamb plates. Only genus Botknocidaris (Fig. 78, a),
Ordovician of Baltic Provinces, the only known Ordo-
vician echinoid.

2. ECHINOCYSTOIDA. Two aberrant forms, morpho-
logically exocyclic, though not related to the Irregulares.
Echinocystis and Palaodiscus, Silurian of Shropshire.

3. MELONECHINOIDA. A series in which both amb
and interamb columns tend to increase in numbers ; no
large tubercles, only granules with small radioles.
Silurian to Permian. Palaechinus (Sil.-Carb., Fig. 78, b),
Melonechinus (Carb., Fig. 78, d).

4. CIDAROIDA. A series in which the ambs have never
more than two columns, but in Palaeozoic genera and
in the Barremian (Tetracidaris) the interambs have more
than two. Even where there are only two interamb
columns the interambs are many times wider than the
ambs. Large tubercles and radioles on the interambs

272

PALEONTOLOGY

only. Amb plates simple . Devonian to Recent. Arckafr
cidaris (Carb.-Perm.), Cidaris (Trias. -Rec., Fig. 78, c, e).

FIG. 78. — REGULAR ECHINOIDS.

a, Rothriocidaris pahlem Schmidt, Ordovician, Russia. (Natural size.)
Interambs shaded, b, Palceechinus elegans M'Coy, Lower Carbo-
niferous. (x£. ) Details of ambs omitted, c, Cidaris florigemma
Phillips, Argovian (Coral Rag). (x£.) Details of ambs omitted.
d, Melonechinus liratus Jackson, Lower Carboniferous. Portion of
one amb and one interamb (full breadth). (X§.) e, Cidaris smithi
Wright, Argovian (Coral Rag). One interamb plate, with perforate
tubercle, and the amb plates of the adjacent single column. Slightly
enlarged. /, Peltasfes wrighti Desor, Aptian, Faringdon (Berks).
Apical disc. (Natural size.) (After Wright.) A, periproct; Mad.,
madreporite. Between these is the pentagonal stir - anal plate.
o-( Cyphosoma konigi Mantell, Emscherian (Upper Chalk). Compound
amb plate. (X3.) Traces of the sutures between the original six
plates are seen. Tubercle imperforate. h, Stomechinus germinans
(Phillips), Aalenian (Pea Grit). Two interamb plates and the adjacent
amb plates. (Xi-) Showing triserial character of ambs, and imper-
forate tubercles, a, d, after R. T. Jackson; b, after Baily; remainder
after Wright.

5. PLESIOCIDAROIDA. — An aberrant Triassic group,
with very large apical disc, and only three plates in the
interambs. Tiarechinus (Carnic.).

THE ECHINODERMATA 273

6. CENTRECHINOIDA [DIADEMOIDA]. — Regular echi-
noids in which there is a tendency to form compound
plates in the ambs, which are not much narrower than
the interambs Peltastes (Jur.-Cret), Salenia (Cret.-Rec.),
and Acrosalenia (Jur.-Cret.). have a large apical disc with
an extra suranal plate (Fig. 78, /). Hemicidaris (Jur. ) has
been described. In Pseitdodiadema (Jur.-Cret.) three
primaries are fused, but the pores remain uniserial except
close to the peristome. In Diplopodia ( Jur.-Cret. j, with
the same type of compound plate, the pores are bisenal
— i.e., they form two vertical columns within each column
of plates; while in Stomechinus (Jur.-Cret., Fig. 78, //)
they are triserial.

In Cyphosowa (Jur.-Cret., Fig. 78, g) as many as six
plates, three primary and three demi-, may fuse into a
compound plate to carry the big tubercles, which are
imperforate (the last three genera having them per-
forate).

Echinus (Cret.-Rec.), the common sea-urchin, has
plates compounded of two primaries and one demi- ;
pores triserial ; poriferous zone narrow ; tubercles and
radicles numerous and not very large.

Pelanechinus (Jur.) and Echinothtiria (Cret.) form a re-
markable group, of which there are living survivors, in
which the whole corona is flexible, the plates being
movably articulated by bevelled edges.

SUB-CLASS : Irregularia (Exocyclica). — Sea-urchins
of this sub-class appear first near the top of the Lower
Jurassic (Aalenian stage), and the simplest of them,
Pygaster (Fig. 79, a), is very little removed from some
simple member of the Regular Centrechinoida, the peri-
proct having shifted to a position only just outside the
apical system. With this, however, are others with much
stronger irregularity (Hyboclypeus), the mouth having also
shifted forwards from its central -position and having

18

274

PALAEONTOLOGY

FIG. 79. — (For description^see p. 275.)

THE ECHINODERMATA 275

lost its jaws and teeth. From the beginning we thus
find two well-marked and divergent orders.

ORDER i : GNATHOSTOMATA. — Mouth central, with
girdle and lantern (the latter sometimes vestigial). All
ambs similar. Symmetry bilateral, but with strong
radial tendency.

Two divergent sub-orders are found : (i) Holectypina,
in which the jaws are reduced in size and strength, the
pores of the ambs extend in unbroken series from apex
to base, and some of the amb-plates may be compound ;
and (2) Clypeastrina, in which the jaws are very powerful,
and the ambs tend to be petaloid, but there are no com-
pound plates.

The chief genera of (i) are Pygaster (Jur.-Cret,
Fig. 79, a), with periproct close behind apical disc ;
Holectypus (Jur.-Cret., Fig. 79, b), in which it has shifted
to the under side ; Discoidea (Cret.), hemispherical, peri-
proct marginal, with ten radiating ridges internally around
mouth, well seen on internal casts ; and Comilus (Cret.),
already described.

In (2) the tendency is for the shape to become very flat
(cake-urchins) : this diminishes the strength of the corona
against vertical pressure, and two methods are adopted
to strengthen it. The more usual is for vertical pillars or

FIG. 79. — IRREGULAR ECHINOIDS.

a, Pygaster morrisi Wright, Bathonian. Aboral view. ( X \. ) a', Apical
disc of P. umbrella Agassiz. (X5) b, Holectypus depressus Leske,
Bathonian. Oral view. (Xj.) b1 ', Apical disc. (x|.) c, Clypeus
altus M'Coy, Vesulian. Aboral view. (X£. ) c1, Oral view. (Xj.)
c", Apical disc. (Xf.) c'", Interamb plate and adjacent part ct
amb. (X2.) d, Clypeaster su/arcinatus Duncan and Sladen. Aboral
view, d' ', Oral view, d" ', Side view. (All X \. ) a, b, c, After Wright ;
d, after Duncan and Sladen. In all, except the apical discs, the
sutures between the plates are omitted ; in a and b, pore-pairs are
indicated by single dots. In the apical discs the madreporite is dotted ;
the ocular plates are distinguished by smaller size and absence of
genital pores ; in b', the left anterior ocular plate and pores on four
genital plates are accidentally omitted. A, periproct; A.F., actinal
furrow; Ap.t apical disc; B, bourrelet ; M, peristome ; Ph., phyllode

276 PALAEONTOLOGY

radiating partitions to be developed internally to support
the roof. The other is for the margin to be notched in the
interambs, the plates in the sides of these notches serving
as supports of the roof : further growth may reunite the
plates at the margin, converting notches into perforations
(" lunules "). In all but the more primitive genera some
of the amb plates widen out and pinch the interambs
across until they meet the plates of the next amb ^dis-
continuous interambs), and there are furrows (actinal
furrows) radiating from the mouth along the ambs.

Chief genera : Echinocyamus (Cret -Rec.) and Fibularia
(Cret.-Rec ) are very small primitive forms with petaloid

FlG. 80. — COLLYRITES RINGENS, AGASSIZ, VESULIAN.

Aboral view. (Xf.) Ptriproct seen at lower end. The four black spots
near centre are the genital pores of the apical disc ; from here the
" trivium " radiates, while the " bivium " is far behind. (After
Wright.)

ambs, but none of the other special faatures of the sub-
order. Scutella (Eoc.-Rec.) is very flat, circular, with
bifurcating actinal furrows ; Mellita and Rotula (Plio.~
Rec.) have lunules; Clypeastev (Fig. 79, d, Olig.-Rec.) is
oval to pentagonal in plan, with straight actinal furrows
in some species advantage is taken of the strengthening
pillars to raise the central part of the roof, including the
respiratory petals, into a dome. Such species are the
largest and most massive of all sea-urchins. All but
the most primitive genera of this sub-order are confined
to warm seas.

ORDER 2 : ATELOSTOMATA. — Mouth shifted forward ;
lantern and jaws lost. Anterior amb somewhat different

THE ECHINODERMATA

277

from the other ambs. Symmetry strongly bilateral.
Tendency to separate the three anterior ambs (trivium)
from the two posterior (bivium).

There are two sub-orders — (i) Cassiduloidea, in which
the mouth, though slightly forward, still opens down-
wards ; and (2) Spatangoidea, in which the mouth opens
forwards, with more or less of an underlip, and between
it and the anus the posterior interamb plates form a
strong convex plastron.

Chief genera of (i) are Nucleolites [Echinobrissus] (Jur.-

FlG. 8l. — ECHINOCORYS SCUTATUS, LESKE, EMSCHERIAN
(UPPER CHALK).

a, Side view; b, apical disc; c, oral view. (Allx J.) (After Wright.)

Cainozoic), with oval outline, broader and rather trun-
cated behind, with deep anal groove on upper surface,
ambs sub-petaloid, the outer pore of each pair elongated ;
Clypeus (Fig. 79, c, Jur.), flat, circular, anal groove as in
the last, ambs petaloid, elongation of outer pores very
great, mouth surrounded by a floscelle — five ambulacral
grooves (phyllodes), with protuberances (bouvrelets) on the
interambs; Echinolampas (Eoc.-Rec.), ovoid, somewhat
conical, ambs petaloid, anus sub-marginal, floscelle
present ; Collyrites (Fig. 80, Jur.-Cret), oval, with rounded
base, with strong separation of bivium and trivium.

278 PALEONTOLOGY

Chief genera of (2) are Echinocorys \_Ananchytes~\ (Fig. 81,
Cret), oval, with flat base, periproct sub-marginal; Holaster
(Fig. 82, Cret.-Mio.), heart-shaped with anterior groove,
slight separation of bivium and trivium ; Micraster (Cret.-
Mio.), already described ; Hemiastev (Cret.-Rec.) and
Schizasttr (Eoc.-Rec.) differ from Micraster in not having
a sub-anal fasciole, but have a peripetalous fascicle running
round the petaloid ambs, which are deeply sunk, especially
in Schizaster, which has also a lateral fasciole branching
off from the main fasciole on each side.

The Echinoidea are by far the most important, geo-
logically, of the Eleutherozoa. Of the Stelleroidea it is

FIG. 82. — HOLASTER SUBGLOBOSUS, LESKE, CENOMANIAN.
ft, Aboral view. ( X £. ) bt Apical disc. (X$.) (After Wright.)

only necessary to say that they are known from the
Cambrian upwards, and where they do occur it is fre-
quently in " starfish-beds" (e.g., where representatives of
one or a few species are extremely abundant), yet between
such beds there are enormous thicknesses of rock desti-
tute of any trace of them. A knowledge of them is,
therefore, far less needed by the geologist than that of
echinoids or crinoids.

The Holothurians have their skeleton reduced to
microscopic spicules, so that special search has to be
made for them. They are found, here and there, from
the Cambrian upwards.

THE ECHINODERMATA 279

Short Bibliography.

BARRANDE, J. — Systeme Silurien de Boheme, vol. vil,
Cystids and Crinoids (1887).

BATHER, F. A.— (i) British Fossil Crinoids, Annals and
Mag. Nat. Hist., ser. 6, vols. v.-ix. (1890-92) ; (2) review
of Wachsmuth and Springer's Crinoid Monograph,
Geol. Mag., dec. 4, vols. v.-vi. (1898-99).

BATHER, F. A., AND GREGORY, J. W.— Echinodermata
in Lankester's Treatise on Zoology, part 3 (1900).

D'ORBIGNY, A., CONTINUED BY COTTEAU, G. H. —
Paleontologie fran9aise [Echinoidea] Terrains juras-
siques, vols. ix., x. ; Terrains cretacees, vols. vi., vii. ;
Terrains tertiaires, vols. i., ii. (1853-94).

DUNCAN, P. M., AND SLADEN, W. P. — Fossil I&hi-
noidea of Western Sind ... of Kach and Kattywar,
Pal. Indica, ser. 14, vol. i. (1882-6).

FORBES, E. — British Silurian Cystidea, Mem. Geol.
Survey, vol. ii., part 2 (1848).

GREGORY, J. W. — Revision of British Cainozoic Echi-
noidea, Pvoc. Geol. Assoc.t vol. xii. (1891).

JACKSON, R. T. — Phylogeny of the Echini, Mem. Boston
Soc. Nat. Hist., vol. vii. (1912).

ROWE, A. W. — Analysis of Genus Micraster, Quart.
Jonrn. Geol. Soc., vol. Iv. (1899).

WACHSMUTH, C., AND SPRINGER, F. — North-American
Crinoidea Camerata, Mem. Mus. Comp. Zool. Harvard,
vols. xx., xxi. (1897).

PAL/EONTOGRAPHICAL SOCIETY'S MONOGRAPHS.

FORBES, E. — Tertiary Echinoderms (1852).

SLADEN, W. P., AND SPENCER, W. K. — Cretaceous
Echinoderms, vol. ii. : Asteroidea and Ophiuroidea
(1891-1908).

280 PALEONTOLOGY

WRIGHT, T.— (i) Oolitic Echinoderms, vol. i. : Echi-
noidea (1857 78); vol. ii. : Asteroidea and Ophiuroidea
(1863-80); (2) Cretaceous Echinoderms, vol. i. : Echi-
noidea (1864-82).

BRITISH MUSEUM CATALOGUES.
ETHERIDGE, R., AND CARPENTER, P. H. — Blastoidea.

VIII
THE GRAPTOLITES AND CORALS

THE Graptolites are an extinct group, found only in the
Older Palaeozoic rocks, and most abundantly preserved
in black shales. Their habit of growth is more sugges-
tive of plants than animals, consisting as they do of
numerous similar parts borne on a stem which may be
branched. Such a habit is, however, known in living
aquatic forms which are undoubtedly animals : it is
easily arrived at when very simple structure is associated
with a fixed habit. So far as this feature goes the
graptolites might be assigned to any one of several
animal phyla, but the balance of opinion is in favour of
regarding them as hydrozoa (see later, p. 296). The
names given to the parts of the graptolite are chosen
on this assumption.

The skeleton of graptolites is composed of some
organic material. When preserved in shales it is usually
crushed flat, and appears as a thin film, sometimes of a
whitish material, more often of graphite. More rarely
the material was replaced by pyrite before being crushed,
and still more rarely it has been preserved uncrushed
and unaltered in limestone from which it can be ex-
tracted by treatment with acid.

281

282 PALEONTOLOGY

i. Didymograptus murchisoni (Fig. 83) is an
abundant graptolite in the black shales of Middle
Ordovician age of Abereiddy Bay, Pembrokeshire (and
elsewhere). Its shape is well described by the name
" tuning fork graptolite." Corresponding to the stem of
the fork is an acutely conical body (crushed flat in the
shale) called the sicula, its apex drawn out into a short
thread called the nema. There is reason to believe that by

FlG. 83. — DlDYMOGRAPTDS MURCHISONI (BECK), UPPER

LLANVIRNIAN.
(Natural size.) (After Elles and Wood.)

S, Sicula.

the nema the graptolite was attached to floating seaweed,
from which it hung downwards, and we will describe the
fossil as if in this position. In this description many state
ments. cannot be fully verified on ordinary specimens, the
facts having been demonstrated by Dr. Holm and other
Swedish palaeontologists on specimens preserved un-
crushed in limestone.

The sicula probably lodged the original polyp, from
which other polyps afterwards were formed by a process

THE GRAPTOLITES AND CORALS 283

of budding (Fig. 84). These are arranged along two
branches, or stipes, hanging down almost vertically but
slightly divergent. Each stipe consists of a series of
somewhat cylindrical cups, or thecce (hydrothecse), which
lodged a polyp each, and a common canal uniting their
inner ends. The sicula and stipes constitute thepolypary,oic
complete skeleton, of the colony of polyps. The end at

FlG. 84. — DlDYMOGRAPTUS MINUTUS, ToRNQUIST, LOWER

ORDOVICIAN (ORTHOCERAS LIMESTONE), OLAND.

Initial portion, reverse side (Xig). (After Holm.) S', Apical part of
sicula; S", aperture of sicula; Cr. C., crossing canal; c,c.t common
canal ; i1, 21, first and second thecae of first stipe ; i2, first sicula of
second stipe.

which is the sicula is called the proximal end, the opposite
end the distal.

The two stipes are not, however, absolutely sym-
metrical. The first theca of one stipe arises directly from
the sicula, and successive thecae, with the common canal
pertaining thereto, are budded off in turn ; but the first
theca of the other stipe is budded off from the first theca
of the first stipe, so that its portion of common canal has

284 PAL/EONTOLOGY

to cross from one side of the sicula to the other. This
is called the crossing canal, and the side on which it lies
is called the reverse, while the side on which the sicula is
in front is the obverse. After thecae have thus been
formed for each stipe, further thecse are developed alike
on each. Each theca is straight, and being at an acute
angle to the direction of the stipe, overlaps the adjacent
theca considerably. If tangents are drawn to the outer
face of the first theca on each side, the angle enclosed by
them is the angle of divergence, which in this species is less
than 45°. This angle is not maintained by the stipes, as
lower down they become sub-parallel : a polypary so dis-
posed is described as dependent. Seven British species of
Didymograptus have this dependent form : they are dis-
tinguished from one another by differences of detail in
the thecae.

There are other species which begin their growth like
D. murchisoni and form two stipes, but with a different
angle of divergence or different final direction.

Thus if the angle of divergence is about 90° and the
stipes continue at that angle, the polypary is said to be
declined ; if starting at less than 90° they gradually
spread out almost horizontally, deflexed ; if with an
angle of divergence less than 180°, they quickly be-
come horizontal ; if curved upwards, reclined; if curved
upwards and then downwards, reflexed (Fig. 85). Even
in these last two cases (known only in one species each)
the original angle of divergence is less than 180°. The
constancy of these angles and curvatures, as well as
the frequent abrupt bends in stipes that have been

5.1

ll

1 1

g w

1 a
Q. a

IS

i§

Z ofi

t -

•2 co
ll

III

II 3

286 PAL/EONTOLOGY

snapped by too great strain (as in Fig. 88, c) shows that
the polypary was a fairly stiff structure.

All these forms, covering twenty-seven British species
alone, are habitually included in the one genus Didymo-
graptus. But the modern conception of a genus is that
of a number of species connected together and separated
from species of other genera by an immediate common
descent. That Didymograptus, in its accepted extent, is a
true genus has been rendered very doubtful by certain
observations made in 1895 by Marr and Nicholson.

The genus Didymograptus is found in the Lower and
Middle Ordovician. In the Lower Ordovician only are
found graptolites which start growth like Didymograptus,
but branch again, forming four or eight stipes, and in the
Tremadoc beds (transitional between Cambrian and
Ordovocian) are still more complexly branched forms.
Marr and Nicholson pointed out that similarities of stipe-
disposition and thecal form can be traced through
the morphological genera Bryograptiis (or Clonograptus or
Loganograptus), Tetvagraptus and Didymograptus in these
successive periods, and they suggested that these indicate
the true lines of descent, while the mimber of stipes
underwent reduction from many to four, and at last to
two, in at least five independent lines of descent (Fig. 85).

The reader may have been struck with the fact, which
has puzzled many geologists, that the earliest graptolites
(Tremadoc) are more complex than their descendants in
the Ordovician. Further, that in individual growth a
Tetragraptus must pass through a Didymograptus stage,
while a Clonograptus must pass through first a Didymo-

THE GRAPTOLITES AND CORALS 287

gvaptus and afterwards a Tetragrapttis stage ; and he imay
ask whether this does not contradict all that has been
deduced from other animal phyla as to the relation of
ontogeny to phylogeny.

To this it must be replied, in the first place, that in the
graptolites we know very little of actual ontogeny. The
development of the sicula is ontogeny ; all that follows
is not the development of an individual, but of a colony,
by budding — astogeny. However, students of other
colonial animals (as the Bryozoa) have shown grounds
for the view that astogeny is also, like ontogeny, a re-
capitulation of phylogeny. Apparently, then, the grapto-
lites contradict this conclusion. But we must not stop
short at the Tremadoc graptolites in tracing their
descent. It is inconceivable that these complexly
branched forms should have been derived abruptly from
non-colonial ancestors. They must have had inter-
mediate ancestors (quite unknown,) in which the branch-
ing gradually became more and more complex. The
known graptolites of this family must illustrate the cata-
genetic stages of branching from an acme in the
Tremadoc epoch.

However the extent of the "genera" so far mentioned
may have to be altered, they appear to form a natural
family — the Dichograptida, of which the five series men-
tioned would be sub-families. The family is distinguished
by the simple structure of its thecae, and by the fact that
the first-formed thecae always grow downwards from the
sicula. It ranges from Tremadocian to Llandeilian.

Another " genus " which should find a place here is

288 PALEONTOLOGY

Phyllogfaptns (Fig. 85), which in a purely morphological
classification has been made into a separate family or
even sub-order, since it has four rows of thecse on one
stipe, while all other graptolites have two or one. It
consists of species closely allied to the reclined Tetra-
grapti, but with the four stipes confluent.

2. Climacograptus wilsoni (Fig. 86) is a graptolite
which occurs in such abundance at the base of the Hartfell
Shales in the Southern Uplands of Scotland that it has
been chosen as a zone-fossil. It has an unbranched stipe
with a double column of thecae (biserial polypary), and may
reach a length of six centimetres or more. When com-
plete, the proximal end shows the sicula, having at its
base a very large vesicle (a specific character, not found
in other Climacograpti), and passing at its apex into
a rod, the virgula. This answers to the nema of Didymo-
graptus, but serves an additional function. The first
theca budded from the sicula begins to grow downwards
as in Didymograptus, but soon makes a sharp bend and
grows upwards, and all subsequent thecae grow in this way.
Consequently the virgula comes to be imbedded between
the two columns of thecae and serves as a support to
them. In many species of Climacograptus (though ap-
parently not in C. ivihoni] the virgula extends for some
distance beyond the most distal thecae, and probably
served to suspend the polypary from floating weed.

The thecae near the proximal end have a simple form,
but higher up they acquire a double right-angled bend,
which gives the polypary the appearance of a double row
of square bodies with narrow spaces (excavations] between.

THE GRAPTOLITES AND CORALS 289

Each square is the distal portion of a theca, with its
aperture horizontal and opening upwards into the exca-
vation above it. The middle portion of each theca is
bent horizontally, and the proximal portion is again
vertical.

Specimens of Clirnacograptus are usually flattened in
the plane which bisects both columns of thecae, and
thus give a profile view; but occasionally they are
flattened in the plane separating the two columns : they

a b

FIG. 86. — CLIMACOGRAPTUS WILSONI, LAPWORTH, LOWER
CARADOCIAN.

a, Initial part of polypary (X i), showing sicula with large vesicle, and six
thecae of each column ; b, distal part of polypary ( x i), showing form
of thecae and "excavations " clearly on the left side. (After Elles and
Wood.)

then show one set of thecae only in full face, and from
its ladder-like appearance this is called the scalarijorm
view.

3. Monograptus priodon (Fig. 87) is a very
abundant graptolite in the lower part of the Wenlock
Shale, though not restricted to that horizon. Its
polypary is unbranched and attains a considerable
length — up to a foot (25 c.m.) and more. The sicula
lies on one side of the proximal end, and from its apex
arises a stout virgula which supports a single column of

19

290

PALEONTOLOGY

thecae on the opposite side to the sicula. These thecae
lie obliquely and overlap one another for about half their
length ; the free portion is bent first outwards and then
downwards, so as to have a hook-like appearance. The
curvature of the theca and the direction of the aperture
are thus the opposite to those of Climacograptus. The
first theca grows upwards from the sicula : the first stage
of downward growth, corresponding to the permanent
direction of Didymograptus, and still preserved in Climaco-
graptus, is skipped altogether in Monograptus.

a b c

FlG. 87. — MONOGRAPTDS PRIODON (BRONN), WENLOCK SHALE.

a, Initial part of polypary, showing sicula, with virgula arising from its
apex ; b, distal thecae, reverse aspect, apertures partly embedded in
rock ; c, distal thecse in low-relief, obverse aspect, showing apertural
margins. ( X f) (After Elles and Wood.)

It is very probable that Monograptus was not a floating
organism, but was fixed to the sea-bottom.

The following classification of the graptolites is that
given in the monograph by Elles, Wood, and Lapworth.
Professor Freeh of Breslau unites the three first families
as Axonolipa (without a virgula), and the rest as
Axonophora (with virgula).

Family DICHOGRAPTID^E. — Uniserial Graptoloidea, with
bilateral polyparies, bearing simple subcylindrical thecas.

THE GRAPTOLITES AND CORALS 291

Branching usually dichotomous, but occasionally ir-?
regular. Primary angle of divergence generally 180° or
less. Tremadocian to Middle Llandeilian.

Chief genera, are those mentioned on Fig. 85. Of
these only Didymogvaptus ranges above the Lower
Llanvirnian (bifidus zone).

Family LEPTOGRAPTID^:. — Uniserial, with slender
flexuous polyparies ; primary angle of divergence ap-
proximately 1 80°; branching usually lateral; thecae
elongated, with slight sigmoid curvature, apertures in-
clined, situated partly within depressions (excavations),
somewhat introverted (turned in wards;, but not introtorted
(twisted inwards). Range in Europe : Middle Llanvir-
nian to Middle Ashgillian, but in Australia a species of
Leptograptus is associated with other graptolites of Skid-
da vian facies.

Chief genera: Leptograptus (Fig. 88, b), without lateral
branching (range of family) ; Nemagraptus (Fig. 88, c)t
S-shaped polypary with numerous lateral branches
(Middle Llandeilian); Pleurograptust*with compound lateral
branches (one zone in Hartfell Shales, Ashgillian).

Family DICRANOGRAPTID^;. — Uni-or uni-biserial, angle
of divergence always exceeding 180°; thecae with strong
sigmoid curvature; apertures horizontal or inclined,
situated within well-defined "excavations" and fre-
quently introverted and introtorted. Llanvirnian to
Ashgillian.

There are two genera : Dicellograptus (Fig. 88, J),
in which the two branches rise up somewhat as
in the reclined dichograptids, and Dicranograptus
(Fig. 88, e), in which they are united at first (as in
Phyllogvaptus or Climacograptus), but afterwards separate.
This family appears to be transitional between Lepto-
graptidae and Diplograptidae, but Dicranograptus in
its life-history shows apparently catagenesis from the

292

PALAEONTOLOGY

FIG. 88. — (For description see p. 293.)

THE GRAPTOLITES AND CORALS 293

higher diplograptid stage and may be compared with
those early nautiloids (Lituites, etc.), which show uncoil-
ing before coiled forms had become well established ;
while Di, sllograptus shows only an intermediate stage
without any anagenesis. The latter has the range of the
family, while the former is Llandeilian and Caradocian
only.

Family DIPLOGRAPTID/E. — Biserial, with straight un-
branched stipes. Thecae tubular, usually in contact for
a large part of their length. Range: Middle Skiddavian
to Upper VaLentian, with acme from Caradocian to
Middle Valentian.

Two genera are usually recognized. CUmacogvaptiis
(Fig. 86) is characterized by the sigmoid curvature of the
thecae, and its «' excavations "; Diplogvaptus (Fig. 88, /, g)
has straight or nearly straight thecae, overlapping and with
aperture everted (turned outwards). The latter has been
subdivided into several sub-genera. Dr. Ruedemann of the
New York Geological Survey has discovered specimens
showing that Diplogvaptus lived in clusters hanging from

FIG. 88. — VARIOUS GRAPTOLITES.

a, Dictyonema flabelliforme (Eichwald), Tremadocian. Part of polypary,
with sicula and nema. (Natural size.) (Full length over 80 mm.,
maximam width 75 mm.) b, Leptograptus Jlaccidus (Hall), Carado-
cian. (XT-)- S, Sicula. c, Nemagraptus gracilis (Hall), Llandeilian.
(Xj.) dt Dicellograptus gurleyi Lapworth, Llandeilian, New York.
(Xf.) e, Dicranograttus nicholsoni Hopkinson, Llandeiliaa-Carado-
cian. (f.) /, Diplograptus (Petalograptus) palmeus van tenuis
(Barrande), Valentian. S, Sicula ; V, virgula. g, Diplograptus
foliaceus M'Coy, Llandeilian, New York. Floating colony. (Natural
size.) (After Ruedemann. ) Pn, Pneumatophore (float) ; G, gonangia
(reproductive vesicles) ; P\, first generation of polyparies ; /*$ second
generation ; V, virgula. h, Monograptiu cyphus Lapworth, Lower
Valentian. Thecse near proximal end. i, M. sandersnni Lapworth,
Valentian. Distal thecae. j, M. vomerinus (Nicholson), Salopian.
Thecae near proximal end ; in full relief, showing torsion of thecal axis.
k, M. lobiferus (M'Coy), Valentian. Distal thecae. /, M. convolutus
(Hisinger), Valentian. m, M, (Rastrites) longispinus Perner, Valen-
tian. (Allxf. unless otherwise stated.) a, e, g, After Ruedemann;
b, d, from Ruedemann, after Lapworth; c, from Ruedemann, after
Hall ; remainder after Elles and Wood.

294 PALEONTOLOGY

a float, and that numerous siculae were developed in
reproductive vesicles under the float and grew out from
these as a second generation of poly paries (Fig. 88, g).

Families GLOSSOGRAPTID^E and RETIOLITID^ are
allied to the last, but show peculiar developments of the
test — firstly, in a change from a membrane of uniform
thickness to one with strengthening ribs, at last becoming
a complex network, in the meshes of which is left a
membrane so delicate that it is lost in the fossil state ;
secondly,, in a great development of projecting spines.
The latter form the special feature of the Glossograptidai
(extending through nearly the whole Ordovician), the
network being most highly developed in Retiolitidce
(almost confined to Silurian).

Family DIMORPHOGRAPTID^E is a small group generally
spoken of as transitional between Diplogvaptidcc and
Monograptidw. They may either be regarded as mono-
graptids which revert to a diplograptid stage, or as
diplograptids in which the formation of the second
column is delayed, as they are uniserial only in the early
part of the polypary and biserial afterwards. The only
genus, Dimorphograptns, is confined to the Lower Valentian,
being contemporaneous with the earliest monograptids.

Family MoNOGRAPTiDyE. — Uniserial and unilateral.
The characteristic Silurian family, ranging from bottom
to top of that system.

The chief genus, Monogvaptus (Figs. 87 ; 88, Ji-m), is
divided by Miss Elles and Miss Wood into the following
seven groups, characterized by differences in the thecse,
and named after a type-species :

1. M. cyphus : thecae simple, straight, overlapping
tubes, with even apertural margins (Middle Valentian
to top Salopian ; Fig. 88, h.)

2. M. sandersoni : thecae with flowing sigmoid curvature
and oblique apertural margins (Lower and Middle
Valentian ; Fig. 88, i).

THE GRAPTOLITES AND CORALS 295

3. M. vomerinus : thecae short tubes with abrupt sig-
moid curvature and some torsion of axis (Middle Valen-
tian to Lower Salopian; Fig. 88,/).

4. M. priodon : thecae with apertural region more or less
isolated and retroverted (Middle Valentian to Upper
Salopian ; Fig. 87).

5. M. lobiferus : thecae with apertural region coiled into
a definite lobe (Middle Valentian to Lower Salopian;
Fig. 88, k).

6. M. convolutiis : thecae triangular or conical, with
reflexed apertural margins (Middle and Upper Valentian •
Fig. 88, /).

7. Usually taken as a distinct genus, Rastvites : thecae
more or less linear and isolate, with reflexed terminations
(Middle and Upper Valentian ; Fig. 88, m).

The polypary of Monographs is usually straight or
slightly curved, but increased curvature resulting finally
in a spiral is found in all groups except the second and
third, and is the invariable case in the last two.

The name Cyrtograptus is applied to monograptids in
which branches grow out from certain thecae, at regular
or irregular intervals ; the main stipe is more or less
spiral. Whether this is justifiably taken as a generic
character may be doubted ; but the species with this
character all occur within a narrow time range (practi-
cally Lower Salopian), and although some of them are
almost identical in other characters with particular species
of Monograptus, they are not contemporary with them.

In addition to these well-defined families, there are a
number of other genera, with abundant irregular branch-
ing, classified loosely as " dendroid graptolites." Some
of these may not be graptolites at all, but plants or
colonial animals of some other phylum. One genus,
however, Dictyonema (Fig. 88, a), has thecae of simple type,
like those of dichograptids. Its abundant branches are

296 PALEONTOLOGY

arranged in the form of a hollow cone, and are connected
at intervals by cross-pieces. This genus ranges from
Cambrian to Devonian, but it is only common in one
well-defined zone in the Lower Tremadoc.

The graptolites have proved of the greatest value as
zone-fossils. Thanks to them alone, Professor Lapworth
was able to unravel the complexities of structure of the
Southern Uplands of Scotland, which had been entirely
misinterpreted before. Apart from detailed zoning they
provide any geologist with the means of recognizing
broad divisions : the complex Dichograptidae are easily
recognized as marking the lowest Ordovician, the Mono-
graptidae as distinctive of the Silurian, and so on. Their
zonal distribution appears to be the same in the Ordo-
vician in Europe and North America, though in the
latter region Silurian graptolites (other than Dendroidea)
are very rare : it is only in Australia that an important
difference in range has been found (see under Lepto-
graptidce, p. 291).

The graptolites have been assumed to be hydrozoa —
that is to say, that the individual animal which occupied
each theca is supposed to have been a hydvoid polyp, all
the polyps in a polypary being united by a living cord in
the common canal. A polyp is essentially a cylindrical
sac with an opening (the mouth) at one end, and around
the mouth a circlet of tentacles; in a hydroid polyp, the
internal cavity is undivided.

There are other fossil hydrozoa besides the graptolites :
they form calcareous skeletons and may be spoken of as
hydrocorallines (using that term in a less exact zoological

THE GRAPTOLITES AND CORALS 297

sense than Moseley's Hydrocorallintf). Chief of these are
the StromatopoYoids which form an important constituent
of some of the Ordovician, Silurian, and Devonian lime-
stones. Their skeleton consists of close-set, wavy, con-
centric laminae traversed by abundant vertical pillars, and
the masses of this structure may attain a considerable
diameter and thickness.

Similar structures are found here and there in later
systems — e.g.,, the spheroidal bodies, about 25 mm. in
diameter, found in the Cambridge Greensand, called
Parkeria. Some fossils of this kind may, however, be
calcareous Algae.

A modification of the hydroid polyp adapted to a nectic
life is the Medusa, or jelly-fish. These most delicate of
living organisms would appear to be the most hopeless
to look for among fossils, yet remains of such have been
preserved in such fine-grained sediments as the litho-
graphic Limestone of Solnhofen and the Middle Cambrian
Shale of Mount Stephen, British Columbia.

A more advanced type of polyp than the hydroid is the
coral polyp, in which the internal cavity comes to be divided
into a central throat, leading down from the mouth, and a
series of peripheral chambers separated by radiating parti-
tions (mesenteries). This characterizes the Actinozoa,
which is one of the classes of the phylum Ccelenterata, the
Hydrozoa being another class. The common sea-
anemone is an example of a polyp of this advanced typa,
but it secretes no hard parts (steveom). Those Actinozoa
which do are spoken of. as corals, a term which has no

298 PALEONTOLOGY

exact taxonomic value, but is a very useful one, especially
in palaeontology.

i. Zaphrentis (Fig. 89) is a genus of which there are
several species common in the Carboniferous Limestone.
Though the following description applies essentially to
Z. konincki, much of it is true of other species.

The shape is that of a drinking-horn — that is, a cone
with its axis more or less curved. An ordinary specimen
is about an inch (25 mm.) in height, and about 12 mm. in
diameter at the broad end. The external surface is
marked with concentric rings, obviously lines of growth,
and in some species there are also fine vertical ribbings.*
The curvature of the cone may be in part the effect of the
conditions of growth : a cone fixed by its apex and grow-
ing larger upwards is in unstable equilibrium. If not
firmly fixed it will topple over towards one side, and,
fixing itself in that position, will become curved in the
effort to grow straight upwards. The direction of curva-
ture is always determined by the internal structure to be
described next.

At the broad (distal) end the cone is hollowed out into
a depression (the calyx) about 6 mm. deep, with steep sides
and a nearly flat floor. A well-marked groove, the fossula,
extends from the centre of this floor to that side which

* In specimens from Tournai, Belgium, otherwise very well
preserved the surface ornament is largely obliterated by chemical
alteration of the substance of the coral. The original aragonite has
been replaced by chalcedonic silica, not uniformly (as is often the
case in other fossils), but in a series of rings, alternately thick and
thin. This is known as " beekite "-structure (see Chap. XL). It is
found also in brachiopods, lamellibranchs, etc., in strata of various
ages.

THE GRAPTOLITES AND CORALS

299

corresponds to the convex curve of the horn (Fig. 89, b).
The sides and floor of the calyx bear a radiating series of
ridges, the septa, which are the most distinctive character
of corals in general. In this species they are over sixty
in number and arranged in an alternation of two series.
One series, the major septa (or entosepta), extend from the
margin of the calyx to the centre of its floor. The

FIG. 89. — ZAPHRENTIS KONINCKI, MILNE-EDWARDS AND HAIME.

Diagrams of vertical and transverse sections. (After Carruthers, simplified. )
a, Vertical section. (X2.) Cardinal and counter septa dotted.
F, fossula. b, Transverse section. ( X 5.) Ht Cardinal ; CS, counter ;
A, alar septa; a-g, metasepta, lettered in order of development. The
short unlettered septa are minor septa.

alternate minor septa (or exosepta) are thinner, less
prominent, and confined to the calyx sides. Although
the septa are, in a general way, radial, the presence of
the fossula makes the whole calyx very distinctly bilateral,
and there cannot be said to be more than an approxima-
tion to radial symmetry. The major septum which
corresponds to the middle line of the fossula is shorter

3oo

PAL/KONTOLOGY

and less prominent than the others : it is called the
cardinal septum. The next major septa on either side are
nearly parallel, converging very slightly downwards on
the calyx-sides, and diverging slightly on the floor, where
they form the boundaries of the fossula. It is not easy
to make out any further characters of the septa in the

CD

FIG. 90. — DEVELOPMENT OF MAJOR SEPTA IN A RUGOSE CORAL.

i, 2, 3, Protosepta; a, b, c, metasepta ; i, cardinal and counter septa;
2, alar septa ; 3, counter-lateral septa. Seven successive transverse
sections, giving ontogenetic stages, are shown. The numbering and
lettering denote the order of development of the septa. (After Car-
ruthers and Duerdeo.)

calyx, but by taking transverse and longitudinal sections
through the coral more may be learned.

If a section were taken very near the apex of a well-
preserved specimen, it is probable (from what is known
in allied corals) that at the very early stage of growth
here preserved, six septa only would be present. Owing
to the destruction of the actual apex in most specimens
it is not possible to prove this, but these six primary

THE GRAPTOLITES AND CORALS

301

septa can easily be recognized in slightly later stages,, and
by means of a series of transverse sections the method
by which additional septa are intercalated can be recog-
nized : this method shows that the symmetry of the coral
is bilateral, not radial (Fig. 90). One of the six primary
septa is the cardinal septum, already noted; another
which is and remains exactly opposite to this is called
the counter-septum. The two on either side of this are
the counter -lateral septa ; the two between them and the
cardinal are the alar septa.

FIG. 91.— SEPTAL DEVELOPMENT IN RUGOSA.

Diagrammatic side views of apex, a, Showing cardinal septum in centre,
metasepta arising on both sides of it ; b, showing alar septum in
centre, cardinal on right margin, counter on left, metasepta arising on
only one side of alar septum, away from cardinal.

Professor Duerden distinguishes the six first septa as
protosepta, and the major septa that arise later as
metasepta. As growth proceeds, additional major septa
appear between the cardinal and alar septa, and between
the alar and counter-lateral, but none between the
counter-lateral and counter-septum (Fig. 91). There are
thus four spaces only in which new septa appear, and the
newest septum in each space is always that nearest the
cardinal septum. The minor septa, on the other hand,
appear simultaneously after the full number of major
septa has been formed.

302 PALEONTOLOGY

The transverse sections also show that the cone is
hollow, consisting of a visceral cavity with a comparatively
thin wall (the thecci), and divided by the septa into loculi.
Near the theca there may be seen a few curved cal-
careous plates crossing some of the loculi : these are
called dissepiments.

A vertical section shows that the floor of the calyx is
the last of a series of horizontal partitions (tabula) across
the visceral cavity, which must be formed, as growth pro-
ceeds, in somewhat the same way as the septa are formed
in a cephalopod-shell. They do not, however, serve the
same function, as all corals are fixed animals; they are
much less regular in arrangement, and have no perfora-
tion like a siphuncle. Each tabula is convex upwards,
and in addition to the cardinal fossula there can some-
times be recognized three other similar, but much
slighter, fossulae, in the position of the two alar and
counter-septa. The counter-lateral septa have no
fossulae. The septa are continuous through the tabulae.
In those species otZaphrentis which have external ribbing,
the ribs alternate with the septa.

The general facts of the development of the septa
stated for Zaphrentis apply to a very large series of
Palaeozoic corals, which constitute the extinct order
Rugosa. By the late Professor Haeckel they were called
the "Tetracoralla,'1 because of the four principal septa and
fossulae, and in distinction from the Mesozoic and modern
" Hexacoralla," in which the septa are arranged by sixes.
But the recent discovery of the early appearance of the
counter-lateral septa makes this numerical distinction
inaccurate.

THE GRAPTOLITES AND CORALS 303

When a Zaphventis grows large it may pass from
conical to cylindrical in shape, and then a change takes
place in the septa : instead of continuing as vertical plates
through the later-formed tabulae, they retain the form seen
in the calyx, of low ridges on the upper surface of the
tabula rising up on the inner face of the theca as low
vertical ridges. Such a condition is described as the
" amplexoid habit," because it attains its fullest develop-
ment in the genus Amplexus. In this form, the tabulae
are. so horizontal and far apart, and the septa reduced to
such feeble ridges, that J. Sowerby, who first described
it a century ago, mistook it for a cephalopod-shell, and,
thinking the resemblance to a coral was accidental, called
the species he described Amplexus coralloides. The same
habit is seen, rather less pronounced, in the genus
Caninia, species of which may attain a gigantic size.
This is distinguished from Zaphventis by the far greater
abundance of dissepiments between the septa for some
distance in from the theca. These two genera occur,
like Zaphrentis, in the Carboniferous Limestone of
England and Belgium.

If we now try to form an idea of the animal of which
the skeleton has just been described, we can only do so
by analogy with living corals. The main living part of
the body of Zaphrentis must have lain within and beyond
the calyx, and perhaps it could shrink up within the
calyx completely when safety required it. In structure
it was essentially a polyp, but more complex than a
hydroid polyp. From the tentacle-surrounded mouth a
short tube projected down into the general cavity, and

304 PALEONTOLOGY

from this tube to the outer wall there radiate a series of
vertical partitions (mesenteries), which also extend down
to the base of the general cavity, which is thus incom-
pletely divided into radiating chambers. It might easily
be supposed that the septa are the skeleton of the
mesenteries, but this is not the case : the mesenteries
occupy the centre of the interseptal loculi ; and where
the septa come the wall and floor of the polyp are, as it
were, indented by them, for the whole skeleton is a
secretion of the external layer of the polyp.

2. Lithostrotion irregulare (Fig. 92) is a coral very
abundant in some of the uppermost beds of the Carbon-
iferous Limestone (zone D2). At Wrington, Somerset,
for instance, it occurs in great masses, red-stained by
iron oxide. It is a compound coral, consisting of a mass
of vertical cylindrical corallites (each secreted by a single
polyp), loosely in contact or slightly separated, and
originating one from another by lateral branching. Only
near the base of the corallum (as the whole compound
skeleton is termed) is branching abundant, but it occurs
occasionally at higher levels. Externally each corallite
is marked by slight horizontal wrinkles, closely set. The
calyx of a full-grown specimen shows twenty-four major
septa and as many minor, the former reaching nearly to
the centre, the latter scarcely projecting through a narrow
marginal zone of dissepiments ; but in young specimens
or cross-sections of the early parts of the colony, eighteen
major only can be found, and there may be fewer. In
the centre of the calyx there projects a laterally
compressed structure, the columella. In a vertical

THE GRAPTOLITES AND CORALS 305

median section of a corallite this is seen to be a vertical
pillar extending the whole length. It is crossed by
numerous very regular tabulae, which are very slightly
convex upwards except near the columella, where each
is lifted up into a cone, and towards the margin, where it
falls steeply into the marginal zone of dissepiments. In
the calyx of an adult corallite the septa appear very

FlG. 92.— LlTHOSTROTION IRREGULARE (PHILLIPS), VlSEAN

(CARBONIFEROUS LIMESTONE).

a, Small portion of a young corallum. (x£.) b, Vertical section of a
corallite. ( X 2.) c, Columella ; P.Z., peripheral zone of dissepiments.
c, Transverse section of a full-grown corallite. (X3-) d, Transverse
section of a young corallite. ( X (j.) H, Cardinal ; G, counter ; A , A ,
alar septa, a, b, c, After Milne-Edwards and Haime ; d, original.

perfectly radial. The compression of the columella,
however, indicates a median plane, and on examining a
cross-section of the younger part of a corallite, a
distinctly bilateral arrangement can be recognized, and
the same proto-septa as in Zaphrentis can be made out
(Fig. 92, d).

In other species of Lithostrotion (e.g., L. martini, which

20

306 PALEONTOLOGY

has twenty-eight major septa) the proto-septa are better
defined : in these species, however, the much greater
abundance of dissepiments tends to confuse the appear-
ance of a cross-section.

Some species of Lithostvotion have cylindrical corallites
in a loose bundle, like L. irregularis ; in others (L.
basaltiforme) they are tightly packed together and become
polygonal (mostly hexagonal) in section.

The essential feature of the genus is the laterally com-
pressed colurnella. Its species are confined to, and very
abundant in, the upper stage (Visean) of the Carbon-
iferous Limestone.

3. Parasmilia centralis is a simple coral found in
the White Chalk of England. It is attached to molluscan
or echinoid shells by a spreading base, from which rises
a short cylindrical peduncular portion, soon expanding
into a conical form which may attain a length of 25 mm.
and a diameter of 12 mm. If further growth takes place,
the form becomes cylindrical ; a total length of 75 mm. or
more may be attained. The calyx is nearly circular, and
shows a prominent columella (which in a section is seen
to be spongy in texture), and forty-eight septa showing
almost perfect radial symmetry. Instead of being in two
series, major and minor, there are four series (cycles) of
unequal length. The first cycle consists of six long septa,
reaching the columella ; the second, of six, slightly
shorter, alternating with the first six ; the third, of twelve,
decidedly shorter, alternating with those of both first and
second cycles ; and the fourth of twenty-four, still shorter,
and alternating with all the others. This arrangement

THE GRAPTOLITES AND CORALS

307

differs altogether from that of Zaphrentis and Lithostrotion,
and as it is generally characteristic of Mesozoic and
later corals, they are often united as Hexacoralla in con-
trast to the Palaeozoic Tetracoralla.

A vertical section shows the visceral chamber to be

FIG. 93. — PARASMILIA CENTRALIS (MANTELL).

a, Young specimen. (Xf-) b, Vertical section. (X|-) Trabecular colu-
mella in centre, black ; dots represent granulations on septa ; thick
marginal lines represent theca in section ; edges of half the septa seen
at the top and (broken section) at the bottom, c, Calyx. (X2.)
Numbers denote septa of the four cycles. (After Milne- Edwards and
Haime. )

free from tabulae and dissepiments ; the surfaces of the
septa bear little tubercles. At the edge of the calyx the
septa project and extend across the theca into continuity
with a series of vertical ridges (cosia] on its outer surface;
These costae show by their varying strength an arrange-

308 PALEONTOLOGY

ment in the same cycles as the septa, except that the first
and second cycles are not distinguishable.

Dr. W. D. Lang, of the British Museum, has shown
that these costae pass through a series of stages in on-
togeny. On the peduncle the costae are low and faint
(they are also only twelve in number, the last cycle of
septa not being yet developed); as growth proceeds they
become stronger, but remain quite smooth. This
anagenesis of costae continues for a distance, and is fol-
lowed by a sudden catagenesis, the costae almost disap-
pearing. A second long period of anagenesis follows, but
this time the costae are not smooth, but have a roughened
or etched appearance. This is as far as the development
goes in P. centralis, though if it continues to grow in
cylindrical form there will be several temporary stoppages
of growth, followed by a rejuvenation — a fresh start being
made from a lower stage of anagenesis than had been
reached before the stoppage.

In other species of Parasmilia, the anagenesis of etched
costae is followed by another sudden catagenesis, and
that by a third period of anagenesis in which the costae
have a granulate appearance. There are thus three
types of costae, and if we recognize three stages in each
(low, medium, and high), altogether nine stages. No
species shows all nine. Those which come earliest in
time show the first four (P. serpentina] or six (P. centralis] ;
later species skip the first three stages (tachygenesis),
and show the next five (P. granulata) or six (P. gravest) ;
still later forms (P. cylindrica, P. mantelli) start straight
away with granulate costae, and show the last three

THE GRAPTOLITES AND CORALS 309

stages only. Rejuvenation may also occur to lengthen
out the life-history, but that is an independent pheno-
menon that may intervene at any stage.

The classification of the corals is in a state of flux at
present, and the following scheme is really out-of-date,
but in the absence of any complete modern scheme it
must be retained provisionally. Also the classification of
corals cannot be separated from that of their non-cal-
careous allies, which are here distinguished by an asterisk
as of little importance or none to the palaeontologist.

CLASS : ACTINOZOA (or Anthozoa).
SUB-CLASS: Zoantharia.

ORDER: ACTINIARIA* (sea-anemones).

ANTIPATHARIA* (with horny skeleton).

MADREPORARIA (with calcareous skeleton : the true
corals).

SUB-ORDER : Rugosa. — These are the Palaeozoic
Madreporaria, often termed Tetracoralla, because character-
ized by four principal septa (cardinal, counter, and two
alar), Mesozoic and later corals being termed Hexacoralla,
because their symmetry is based on six main septa.
(The recognition of the early development of the counter-
lateral septa, making six, in the Rugosa has weakened,
but not destroyed, this distinction.) The usual abundance
of tabulae is another general point of distinction in the
Rugosa.

GROUP : Inexpleta. — This includes a few simple
corals, in which the interseptal loculi are almost entirely
undivided, tabulae and dissepiments being very rare.
Common genera : TurUnolopsis, conical, with occasional
tabulae, found as casts in Llandovery Sandstone ; Palceo-
cyclus (Fig. 94, a), of which the commonest species is
discoidal — i.e., the angle of the cone is increased to 180°

310 PALEONTOLOGY

and the theca no longer encloses the visceral chamber,
common in the Wenlock Limestone ; Cyathaxonia, acutely
conical, with strong columella, Carboniferous Limestone.

GROUP : Zaphrentoidea. — Simple corals ; septa and
tabulae well developed, but few dissepiments. Includes
Zaphrattis, Caninia, and Ampltxus, already mentioned.
Streptelama (Ord.-Sil.) is usually associated with them,
but although earlier in time it has a more elaborate
structure.

GROUP: Cyathophylloidea.— Simple or compound,
with a marginal zone of vesicular tissue ; tabulae usually
less extensive than in Zaphrentoidea.

Family CyathophyUida. — Septa with smooth surfaces,
the primary lamella of each septum being thickened
uniformly on each face with a layer of stereoplasma ;
primary septa not strongly marked, symmetry approach-
ing radial. A large family, with chief genera : Cyatho-
phyllum (Sil.-Carb., Fig. 94, c), simple or compound, with
major septa reaching almost to centre, and minor septa
not very much shorter; Acervularia, compound, massive,
the septa swelling out at a certain distance from the
centre so as to unite to form a circular "inner wall";
Lithostrotion, already described ; Omphyma, simple, top-
shaped, with root-like outgrowths, the four main septa
often very conspicuous.

Family Heliophyllidcz. — Septa with stereoplasma
deposited in vertical ridges, otherwise much like Cyatho-
pyllidce. Genera: Heliophyllnm (Sil.-Dev.), simple;
Phillipsastrcea (Dev.-Carb.), compound.

Family Clisiophyllida . — With a large " false columella "
(not a solid rod, but made up of twisted lamellae, having
a "spider's web" appearance in transverse section, and
making a large protuberance in the calyx). Genera :
Clisiophyllum (Sil.-Carb.), simple; Lonsdaleia (Carb.,
Visean, Fig. 94, b), compound ; Dibunophyllum (Visean),

THE GRAPTOLITES AND CORALS 311

simple, with a lamina from cardinal to counter-septum
running through the false columella.

GROUP : Cystiphylloidea. — Simple ; with great de-
velopment of dissepiments, replacing tabulae and leading
to degeneration of the septa. Includes the only corals
with an operculum to the calyx. Genera : Cystiphyllum
(Sil.-Dev.), cylindro-conical, no operculum ; Goniophyllum
(Sil.), four-sided pyramidal, with an operculum of four
pieces; Calceola (Dev., Fig. 94, ^), slipper shaped, with
an operculum of one piece.

The post- Palaeozoic Madreporaria are sometimes
united as Hexacoralla. The order of development of the
septa in them differs greatly from that of Rugosa. They
are usually divided into three orders (or sub-orders),
though it is now very doubtful if these express natural
affinities.

SUB-ORDERS : Aporosa. — Theca and septa compact.
Include two principal families.

Family Tuvbinolidce. — Without dissepiments. Simple.
Genera : TiwUnolia (Eoc.-Rec.), conical, with projecting
lamella and very prominent costae ; Smilotrochus, similar
but no columella ; Fldbellum (Eoc.-Rec.), compressed,
calyx elliptical, costae weak.

Family Astvceidce. — With dissepiments; simple or
compound. Includes the chief reef-building corals.
Genera : (i) Simple — Parasmilia (Cret), already described;
Montlivaltia^ (Jur.-Cainozoic), conical, turbinate or dis-
coidal, no columella, dissepiments abundant, covered
externally by a perishable calcareous skin (epithccd).
(2) Compound — Thecosmilia (Jur.-Cain., Fig. 94, ^),
essentially a Montlivaltia, which becomes compound by
equal fission ; Isastraa (Trias.-Cret, Fig. 94, /), massive,
with prismatic corallites ; Holocystis (Lower Cret., Fig.
94, g), massive, remarkable for superficial resemblance
to some Rugosa, there being four septa longer than the
others, and tabulae being well developed.

PALEONTOLOGY

FIG. 94. — (For description see p. 313.)

THE GRAPTOLITES AND CORALS 3*3

SUB-ORDER: Fungida. — Dissepiments replaced by cal-
careous rods (synapticuke] crossing the interseptal loculi.
Genera: Anabacia (Jur.), simple, discoidal ; Cyclolites
(Jur.-Eoc.), similar, but septa are thin and perforate, thus
approaching the next sub-order ; Thamnastvcea (Jur.),
compound, massive,' septa continuous from one corallite
to its neighbours.

SUB-ORDER: Perforata.— All parts of the skeleton
traversed by pores, which may reduce it to a spongy
network. Genera : Stephanophyllia (Cret.-Rec.), simple,
discoidal ; Dendwphyllia (Eoc.-Rec.), compound, tree-like
(dendroid) ; Goniopova [Litharcea] (Eoc.-Mio.), compound,
massive.

SUB-CLASS : Alcyonaria.— Skeletons of very variable
character. Many extinct forms are doubtfully placed
here. Septa usually feeble or absent, and never showing
either a tetracorallan or hexacorallan plan ; tabulae
well developed. All compound. Genera : Favosites
(Sil.-Carb., Fig. 94, k), massive, corallites polygonal,
septa vestigial, tabulae horizontal and regularly spaced,

FIG. 94. — FOSSIL CORALS.

Palaocyclus porpita (Linn6), Wenlock Limestone. Plan of calyx,
detail omitted from three quadrants, a', Side view. (Xv) b> Lons-
daleiafloriformis (Fleming), Visean (Carboniferous Limestone). Cross-
section of single corallite. (X2.) £', Side of young corallum. (Xj.)
c, Cyathophyllum hexagonum Goldfuss, Middle Devonian. Cross-
section of single corallite. ( X §.) Dissepiments omitted from greater
part. d. Calceola sandalina Lamarck, Middle Devonian. (Xj.)
d' , Operculum. (After Goldfuss. ) e, Thecosmilia annularis (Fleming),
Argovian (Coral Rag). (Xj.) /, Isastrcea oblonga (Fleming), Port-
landian Section of silicified corallum. (Xs-) Thecae and septa
white, g, Holocystis elegans (Fitton), Aptian (Atherfield Clay).
Surface of part of corallum. (X6.) Septa and columella, black;
theca, dotted, h, Heliolites porosa (Goldfuss), Middle Devonian.
Transverse section. (Xl-) i, Haly sites catenularia (Linne"), Silurian.
Diagrammatic view of portion of a corallum. (Xf.) j, Syringopora
ramulosa Goldfuss, Lower Carboniferous. (Natural size.) k, Favo-
sites gothlandica Lamarck, Wenlock Limestone. Diagrammatic view
of four corallites. (X4-) The exterior of three is seen, with mural
pores, the fourth is split vertically in half, showing tabulae, d, After
Goldfuss ; i and k, original ; the rest after Edwards and Haime.

3*4 PALEONTOLOGY

mural pores perforating the thecae; Pachypora (Dev.),
similar but branching, with thickened walls; Alveolites
(Sil.-Dev.), similar to the last, but calyces cuspate-
triangular; Michelinia (Carb.), with tabulae strongly
arched, corallum with thick epitheca bearing root-like
out-growths ; Syringopora (Sil.-Carb., Fig. 94, /), coral-
lites cylindrical, spaced out and connected by horizontal
tubes, tabulae funnel-like; Aulopora (Sil.-Carb.), an en-
crusting network of tubular corallites, with raised trumpet-
like apertures; Halysites (Ord.-Sil., Fig. 94, j), like the
last, but corallites rise up vertically, forming linear
series— the " chain-coral" ; Heliolites (Sil.-Dev., Fig.
94, h), massive, two kinds of corallites, abundant small
polygonal and scattered larger cylindrical, both with
tabulae, the latter with apparent septa.

Short Bibliography.

GRAPTOLITES.

ELLES, G. L., WOOD, E. M. R., AND LAPWORTH, C.—
British Graptolites, Palceontographical Society (1901-18).

RUEDEMANN, R. — Graptolites of New York, 2 vols.
New York State Museum (1904-08).

MARK, ]. E. AND NICHOLSON, H. A. — Phylogeny of
Graptolites, Geol. Mag., dec. 4, vol. ii. (1895).

HOLM, G. — On Didymograptus, Tetvagraptus and Phyllo-
graptus, transl. Elles and Wood, Geol. Mag., dec. 4,
vol. ii. (1895).

The first of the above-mentioned works contains a
detailed historical account and bibliography of the
Graptolites.

STROMATOPOROIDS.

NICHOLSON, H. A. — British Stromatoporoids, Pal.
Soc., 1886-92.

THE GRAPTOLITES AND CORALS 3^5

CORALS.

CARRUTHERS, R. G. — Primary Septal Plan of Rugosa,
Ann. and Mag. Nat. Hist., ser. 7, vol. xviii. (1906).

CARRUTHERS, R. G. — Revision of some Carboniferous
Corals, Geol. Mag., dec. 5, vol. v. (1908).

DUERDEN, J. E. — Morphology of the Madreporaria,
Ann. and Mag. Nat. Hist., ser. 7, vols. x., xi., xvii., xviii,
(1902-06).

DUNCAN, P. M. — British Fossil Corals (2nd Series),
Pal Soc., 1866-72.

LANG, W. D. — Growth Stages in Parasmilia, Proc.
Zool. Soc., London, (1909), p. 285.

MILNE-EDWARDS, H., AND HAIME, J. — British Fossil
Corals, Pal. -Soc. 1850-54.

TOMES, R. F. — Various papers in Quart. Jonrn. Geol.
Soc., vols. xxxiv-xlix. (1878-93).

VAUGHAN, A. — The Carboniferous Limestone Series
(Avonian) of the Avon Gorge, Proc. Bristol Nat. Soc.,
ser. 4, vol. i. (1906).

VAUGHAN, A., AND OTHER AUTHORS. — Various papers
on zonal distribution of Carboniferous Corals in Quart.
Jonrn. Geol. Soc., from 1905 on.

IX

THE PORIFERA AND PROTOZOA

BELOW the Ccelenterata there are two grades of animal
structure, both of importance in palaeontology — the
Porifera or Sponge's, and the Protozoa, simplest of all
animals.

The Porifera are so named because they take in their
food, not by a single mouth as do all animals from
Crelenterates upwards, but by innumerable pores scattered
over the surface of the body. As a consequence of this,
and of their fixed mode of life, the form of the body
is much more variable than in any phylum yet considered,
and is sometimes very indefinite indeed. The skeleton
in the more primitive forms is composed of great numbers
of separate spicules usually of amorphous silica (opal),
but of calcite in one class. These spicules form a sup-
port to all parts of the body, but are only united by soft
tissues, so that on the death of the sponge they fall apart
and are scattered on the sea-floor. In many cases two
kinds of spicules may exist in the same sponge — body
spicules, elongated and often branched, and dermal spicules,
flat or rounded. A simple spicule in the fresh state, when
seen under the microscope, has the appearance of a
fragment of therrnometer-tube, being glassy and with a

THE PORIFERA AND PROTOZOA 317

fine central canal. Some of the most characteristic
forms of spicule are, shown in Fig. 95. Loose siliceous
spicules of this kind accumulate in such amounts on the
sea-floor to-day as to form a "spicular ooze," as, for
instance, off Kerguelen, at a depth of 120 fathoms.
Similar deposits among the stratified rocks have passed

FIG. 95. — PORIFERA.

a, Monactinellid (x 15), and b, Tetractinellid (x 10) spicules from interior
of hollow flint, c, Dermal spicule of a Lithistid, Upper Chalk. ( X 32.)
d, Siphonia tulipa Zittel, Cenomanian (Upper Greensand), War-
minster (Wilts). Vertical median section (stalk broken off) showing
vertical canals opening into central cloaca, e, Fragment of hexac-
tinellid meshwork, showing central canals (black), Neocomian. (X 30.)
ft g, Ventriculites infundibulifonnis S. P. Woodward, Upper Chalk.
/, Side view of imperfect specimen ; gt cross-section. (Both X}.) (All
after Hinde.)

into cherts, the spicules becoming cemented by a secondary
deposit of chalcedony into a very hard rock. Such
sponge-cherts are found in Britain in the Carboniferous
Limestone, at several Jurassic horizons, and particularly
in the Lower and Upper Greensands of the Cretaceous,
In the higher orders of sponges the spicules are united

318 PAL/EONTOLOGY

into a meshwork, and in such cases the form of the body
is preserved in the fossil. The substance never persists
as unaltered opal ; it may simply change into chalcedony,
or may be replaced by pyrites, marcasite (passing into
limonite), or calcite.

The common bath-sponge belongs to a class having a
continuous skeleton of a horny organic substance, but
such forms are not preserved as fossils, and palaeontologists
are only concerned with two classes, distinguished by. the
chemical composition of the skeleton.

CLASS : SILICISPONGLE.*— Skeleton of silica.

ORDER i : Monaxonida.— Spicules one-rayed (Fig.
95, a). Only known fossil by loose spicules, except one
genus Cliona, a boring sponge in lamellibranch shells
and belemnite-guards, known by its borings, often
preserved as internal casts in flint.

ORDER 2 : Tetraxonida. — Spicules four-rayed (Fig.
95 b, c).

SUB-ORDER : CHORISTIDA. — Spicules not united.

SUB-ORDER : LITHISTIDA. — Spicules united into a con-
tinuous skeleton (Fig. 95, c). Genera: Siphonia (Cret.,
Fig. 95, d), pear-shaped, stalked; Doryderma (Cret.),
cylindrical, branched ; Verruculina (Cret. ), cup-shaped or
irregular, with short stalk.

ORDER 3 : Hexactinellida.— Spicules of six rays,
intersecting at right angles (Fig. 95, e). In the earliest
genus Pvotospongia (Camb.) the spicules were very feebly
united, as they are found only a few together, the form
of the body being unknown. The same is the case with

* Zoologists do not admit this as a natural group: they make
Hexactinellida a class in itself, and unite the remainder with the
horny sponges as a class, Demospongiae.

THE PORIFERA AND PROTOZOA 3*9

other Palaeozoic genera, but from the Trias onwards a
continuous skeleton is found. Genera : Ventriculites (Cret.,
Fig. 95, /, g), more or less inverted conical, with roots,
wall of cone folded ; Plocoscyphia (Cret.), frilled laminae
uniting into a coarse net-work ; Cceloptychium (Cret.),
mushroom-shaped, with folded wall. Siliceous sponges
attain their greatest abundance in Europe in the Creta-
ceous system.

CLASS : CALCISPONGI^E. — With calcareous
skeleton. There are several orders, with three-rayed
spicules, not forming a continuous skeleton, but little is
known of them in the fossil state.

ORDER : Pharetrones.— Spicules united into a fibrous
meshwork. Genera : Bavroisia (Cret.),, clusters of double
cylinders, the space between the inner and outer cylinder
being crossed by partitions much coarser than the tabulae
of a coral ; Peronidella (Trias. -Cret.), hollow, cylindrical ;
Raphidonema (Cret.), more or less cup-shaped. Calcareous
sponges are locally very abundant in the Upper Jurassic
of South Germany, the Lower Cretaceous of Faringdon
in Wiltshire and Upware in Cambridgeshire, and the
Upper Cretaceous (Cenomanian) of Westphalia, but are
rare in general.

The Protozoa are animals composed of a single cell,
or a small number of undifferentiated cells. Only two
orders leave fossil remains — the Radiolaria, which form
a beautiful lattice-work skeleton of silica, and the
Foraminifera, which have one of calcite, aragonite, or of
agglutinated foreign bodies. The majority of forms in
both orders are very small, ranging from i mm. diameter
downwards, but some of them occur in such abundance
at certain horizons as to be quite important rock-forming

320 PALEONTOLOGY

organisms. A few genera of Foraminifera attain a much
greater size, up to a maximum of perhaps 80 mm., and
these giant forms are of value as indices of age, though
unfortunately their geographical range is limited.

The Radiolaria range from possibly pre-Cambrian to
the present day : they are divided into two main sub-
orders— Spumellaria, spherical or discoidal, with the range
of the order, and Nassellaria, of various forms, but always
with dissimilar ends, dating from the Devonian only.
Beds of Radiolarian chert are known in the Ordovician
of South Scotland, the Devonian of New South Wales,
the Carboniferous of Devon and Cornwall, and the
Mesozoic of the Alps. These beds have very generally
been taken as analogous to the modern Radiolarian ooze
of the greatest depths of the oceans ; but in the case of
the British Lower Carboniferous cherts their shallow-
water character seems now decisively proved. On the
other hand, the Miocene Radiolarian earths (not con-
solidated into chert) of Barbados and other West Indian
islands are probably of deep-sea origin.

The Foraminifera form a shell which is nearly always
chambered, the chambers being arranged sometimes in a
straight line, sometimes in a zig-zag, but most frequently
in some sort of spiral. They were at one time taken for
cephalopods, but the resemblance in the shells of the
two groups is only homceomorphy of the most general
sort. There are no gas-chambers in a foraminifer, the
cell-protoplasm filling all the chambers as well as ex-
tending to a greater or less extent outside the shell ; this
it does through a well-defined aperture (much smaller

THE PORIFERA AND PROTOZOA

321

than the diameter of the chamber) in the last-formed
chamber, and (in one sub-order) through numerous
perforations of the whole surface. Moreover, the fora-
minifer does not cut off new chambers by secreting
septa — each " septum " formed the terminal part of the
outer shell before it was converted into an internal
structure by the building of a new chamber in front of it.
The ontogeny reveals in many cases a striking dimorphism :
in the same species growth may begin by a very

FIG. 96. — DIMORPHISM IN A FORAMINIFER, BILOCULINA
BULLOIDES, D'ORBIGNY, RECENT.

a. Exterior showing aperture with T-shaped partition. (X^--) bt Cross-
section of microspheric and c, of megalospheric form. ( X 125.) Only
the early chambers are shown, up to the attainment of the regular
alternation of chambers. Chambers numbered in order of develop-
ment. (After Schlumberger. )

small chamber (microspheric form) or by a much larger
one (megalospheric form). The study of living forms
has shown that this is due to an alternation of genera-
tions.

The microspheric form is produced by a sexual pro-
cess, and its ontogeny is complete ; the megalospheric is
produced asexually, and the early stages are skipped
altogether, just as they are in a rejuvenated coral (Fig.
96, b, c). The microspheric form in some species attains

21

322 PAL/EONTOLOGY

a much larger size than the megalospheric ; in others
there is no difference in size. Both forms occur together
in the same strata, the megalospheric forms being more
abundant than the microspheric — a natural result of the
fact that a number of megalospheric generations intervene
between one microspheric and the next. The late Robert
Douville noticed that in the last beds in which a given
species occurred, megalospheric individuals might be
exclusively present— an indication of near approach to
extinction.

The Foraminifera have been divided into three orders,
based on the structure and composition of the test :
(i) Perforata or Hyalina, with test of calcite, transparent
(at least in the more primitive forms) and with abundant
small perforations in addition to (rarely without) a main
aperture in the last formed chamber ; (2) Imperforata or
Porcellanea, with shell of aragonite, less translucent and
of a brown colour by transmitted light, and without per-
forations ; (3) Arenacea, in which the test is formed of
foreign bodies, such as sand-grains, sponge-spicules, or
the small tests of other Foraminifera, held together by an
organic secretion. The first two orders are well defined,
but the third is probably a rather mixed assemblage of
forms which have been led by special circumstances to
adopt this curious way of protecting themselves in place
of secreting a shell.

The five principal families of the Perforata are as
follows :

i. Lagenidae, the simplest forms, with thin trans-
parent shell. Genera : Lagena, single-chambered, flask-
shaped ; Nodosaria, a straight row of chambers (ortho-
cone) ; CrisUllaria, spiral (ophiocone to sphaerocone).
These all appear to range from Lower Cambrian to

THE PORIFERA AND PROTOZOA 323

Recent; Webbina (Vitriwebbina), an irregular row of
elliptical chambers, encrusting shells (Jur.-Rec.).

2. Textularidae, with a zig - zag alternation of
chambers. Chief genus : Textularia (Camb.-Rec.).

3. Globigerinidse, with chambers globose, usually in
a somewhat irregular spiral, perforations coarse. Chief
genus : Globigerina (Camb.-Rec.), a most abundant
pelagic form at the present day. Associated with it is
the perfectly spherical Orbulina, of which most specimens
when broken open are found quite hollow, but occa-
sionally a Globigerina is found inside : this suggests that
Orbulina is the final stage of Globigerina, the majority of
specimens being megalospheric and omitting the Glo-
bigerina-stage in ontogeny.

4. Rotalidas, with chambers coiled in an asymmetric
spiral, all the whorls being visible on one face, and only
the last whorl on the other. In this family first appears
the supplemental skeleton, an outer calcareous layer plastered
over the surface of the ordinary perforate wall, pene-
trating between the two folds of the latter which here
form the septa, and sometimes forming a surface orna-
ment (even long spines in some) which quite obscures the
arrangement of the chambers.

Chief genera : Discorbina (Cret.-Rec.) ; Planovbulina
(Carb.-Rec.) ; Truncatit,lina(Ca.rb.-Rec.);Anomalina(Cret.-
Rec.) ; Pulvinulina (Jur.-Rec.), with the form described
above, but differing in the relative convexity of the two
faces, without supplemental skeleton ; Rotalia (Jur.-Rec.),
with supplemental skeleton, but recognizably spiral ;
Calcarina, discoidal with peripheral spines, spiral inter-
nally (Cret.-Rec., chiefly in Maestrichtian) ; Tinoporus,
similar shape but chambers irregularly arranged, and
surface with pattern of tubercles (Eoc.-Rec., abundant
on modern coral-reefs).

324 PALAEONTOLOGY

5. Nummulinidae, spiral or cyclical, with subdivided
chambers, and great development of supplemental
skeleton, surface generally smooth, never spinose. This
family includes most of the giant Foraminifera, and most
of those useful as time-indices. Genera : Fusulina (Carb.-
Perm.), spindle-shaped, spirally coiled about the long
axis ; Polystomella (Cret.-Rec.), lenticular, last whorl only
visible externally, its chambers and their subdivisions
recognizable on the surface ; Nummulites or Nummulina
(Carb. ? Eoc.-Rec.), discoidal or lenticular (Fig. 97) ;

d

FIG. 97. — NUMMULITES.

rt, A^. vicaryi d'Archiac and Haime. Edge view. (Xj.) b, N. com-
planatus Lamarck. Edge view, (xj.) c, rf, A7, garansensis Joly et
Leymerie. c, Part of horizontal section. (X8.) d, Part of vertical
section. ( X 12.) (All after d'Archiac and Haime.)

Assilina(~Eoc.), similar, but the later whorls only partially
overlap the earlier. These and other genera are spiral,
but in one sub-family, generally known as the Orbitoides,
the spiral mode of growth is abandoned (except in the
early stages of ontogeny in microspheric individuals) for
a cyclical growth, chambers being added in concentric
rings, which more or less overlap the sides : a medium
horizontal section shows the structure in its simplest
form, and enables the genera to be distinguished easily.
Thus the " equatorial " chamberlets of Orbitoides (Cret.,
Upper Campanian) are rounded ; those of Orthophragmina

THE PORIFERA AND PROTOZOA

325

(Eoc,, Fig. 98) rectangular ; of Lepidocyclina (Oligo.,
Fig. 99) more or less hexagonal ; while in Miogypsina
(Mio.) there is a more distinct spiral "nucleus." The
generic distinctions given above are those most readily
made out in thin sections of Limestones, but are not the
only ones.

The Imperforata are regarded as all belonging to one
family, Miliolidae. The simplest example is Biloculina
(Trias.-Rec., Fig. 96), in which the adult chambers are
hemispheroidal, the aperture with a T-shaped central
septum, alternately at one end and the other as new

FIG. 98. — ORTHOPHRAGMINA UMBILICATA, DEPRAT, EOCENE,
NEW CALEDONIA.

a, Vertical section. ( X !•) b. Portion of a horizontal section.
(After Deprat. )

(Xio.)

chambers are added. The early chambers of the micro-
spheric form show an arrangement found in the adults
of other genera, as Spivoloculina and Miliolina. The
family also includes some large, though scarcely giant,
forms — A Iveolina is a homoeomorph of Fusitlina, but very
different in age (Cret.-Rec., maximum in Eocene), and
Orbitolites (Eoc.), and Marginopora (Cret.— Rec.) resembles
the large discoidal genera of the Nummulinidse : the
chambers, except in the initial stages, are arranged in
cycles with a very regular " engine-turned " pattern,
visible externally owing to absence of overlapping. This

326

PAL/EONTOLOGY

last genus is very abundantly represented on modern
coral beaches.

Of the Arenacea the most important are Saccammina
(Carb.-Rec.) of one or several spheroidal chambers ;
Endothyra (Carb.-Trias.) resembling some of the Rotalidae
in form ; and Orbitolina (Cret.) conical, the chambers
cyclical in arrangement. Textularia is commonly in-

FIG. 99.— LEPIDOCYCLINA TOURNOUERI, DOUVILLE AND LEMOINE,

OLTGOCENE.
(Megalospheric form.)

a, Vertical section. ( X 13.) b, Part of horizontal section. ( X 18.)
(After R. Douvilld and Lemoine.)

eluded among Arenacea, but is always perforate and
hyaline, at least in early life.

The Perforata are usually well-preserved fossils,
whether in Limestone or Clay, or even in sands in some
of the latest formations. The Imperforata, having
aragonite tests, are less frequently preserved. A very
special mode of preservation is that of internal casts in

THE PORIF'ERA AND PROTOZOA 327

glauconite. The formation of these goes on at the
present day on certain areas of the ocean floor. By
some means not yet fully explained this complex hydrated
silicate of iron, aluminium, and potassium is precipitated
in the interior of these minute tests, which themselves
may afterwards undergo solution. The resulting casts
form "green sands," or on a muddy bottom "green
muds." Similar deposits are found at many geological
horizons, but mosf abundantly in the Cambrian and
Cretaceous systems.

As rock-forming organisms the Foraminifera are
important. A local example is the " spotted post " in the
Carboniferous Limestone of the north of England, full
of the arenaceous Saccammina cavteri. Of far greater
importance are the Fusulina-limestones of the Upper
Carboniferous, found in the Eastern Alps, Russia, China
and Japan, the East Indies, and the southern United
States. The arenaceous Orbitolina [Patellina] gives its
name to a limestone in the Aptian of Switzerland. The
White Chalk is made up in great measure of the
remains of Globigerina and Textulavia. The Eocene
Calcaire grossier of the Paris basin (the Paris building-
stone) is full of Miliolidae, and an Alveolina-limestone
occurs in the Hampshire Basin. Far more important
are the Eocene Nummulitic limestones, found throughout
the area of the great ocean (Tethys) of that period, i.e.
in the Atlas, Pyrenees, Alps, Carpathians, Balkans,
Egypt, Persia, the Himalayas, the East Indies and New
Caledonia. In many of these limestones Orthophragmina
is associated with Nummulites. The Oligocene Lepido-

328 PAL/EONTOLOGY

cyclina limestones (commonly known as orbitoidal) have
a similar distribution and are found in the West Indies
also. Although the accumulation of these giant Forami-
nifera to a sufficient extent to form masses of lime-
stone is confined to the site of this great tropical and
sub-tropical ocean, the genera are occasionally found
beyond its limits. Thus Nummulites spread as far as
Hampshire in the latter part of the Eocene period, and
Lepidocyclina into Bavaria in the Oligocene.

This completes the survey of the Animal Kingdom
as known in the fossil state, but for a few minor groups
that have been omitted and may briefly be mentioned
here.

The Bryozoa (or Polyzoa) are a phylum or sub-
phylum of fixed compound animals, having some super-
ficial likeness to the corals, but belong to a decidedly
higher grade. They are often united with the brachiopods
under the name Molluscoidea. Their skeleton (zoarium} is
made up of many more or less tubular individual
skeletons (zocecia), which are much smaller in size than
most corallites, and have no septa. In a few simple
cases the zoarium is like a branching thread ; very often
it forms a flat lamina, either spreading freely in the
water or encrusting other organisms; sometimes it is
massive, in which case it may become difficult to draw
the line between this phylum and the Alcyonaria,
especially as structures analogous to tabulae may be
present. Certain Palaeozoic forms have been referred to
the one phylum by some investigators, and to the other
by others. The Bryozoa are almost exclusively marine,
and range from the Ordovician to the Recent period.
Occasionally they occur in such abundance as to form

THE PORIFERA AND PROTOZOA 329

bryozoal deposits, for instance the Coralline Crag (Lower
Pliocene) of East Anglia.

The Annelida (segmented worms) are for the most
soft-bodied and rarely preserved ; but one order, Tubicola,
is characterized by a fixed habit and the secretion of a
tube. These tubes may be nearly straight, regularly
curved, spiral or irregularly twisted. The curved tubes
might be confused with Dentaliiun, the irregular for an
uncoiled, and the spiral for an ordinary gastropod, and
there are cases where the proper placing of such a fossil
is still doubtful ; but in general there is a roughness and
want of regularity about annelid tubes that prevents
confusion between them and molluscan tests.

Short Bibliography.

PORIFERA.

HINDE, G. J. — (i) British Museum Catalogue of Fossil
Sponges (1883); (2) British Fossil Sponges, Pal. Soc.
(1887-1912) ; (3) On Beds of Sponge-Remains in the
Greensands of the South of England, Phil. Trans. (1885).

MINCHIN, E. A. — Sponges, in Lankester's Treatise on
Zoology, Part II.

ZITTEL, K. A. — Studies on Fossil Sponges. Trans-
lated by W. A. Dallas, Ann. Nat. Mag. Hist. (1877-79).

PROTOZOA.

BRADY, H. B. — Challenger Reports, vol. ix. (Zoology) :
Foraminifera (1884).

CARPENTER, W. B., PARKER, W. K., AND JONES, T. R.
— Introduction to the Study of the Foraminifera, Ray Soc.
(1862).

CHAPMAN, F. — The Foraminifera (1902). Contains a
good bibliography.

330 PALEONTOLOGY

DOUVILLE, R. AND LEMOiNE, P. — Lepidocyclines, Mem.
Soc. Geol. France (Paleontologie), vol. xii , no. 32 (1904)

DOUVILLE, H. — Numerous papers in Bull. Soc. Geol.
France (1902 onwards).

SCHLUMBERGER, C. — Dimorphism of Foraminifera,
Ann. Mag. Nat. Hist. (5) vol. xi. (1883); and papers in
Bull. Soc. Geol. France (1884-1900).

SHERBORN, C. D. — (i) Bibliography of Foraminifera,
recent and fossil, 1565-1888 (1888). (2) Index to genera
and species of Foraminifera, Smithsonian Misc. Coll.,
vol. xxxvii. (1893-96).

VREDENBURG, E. — (i) Zonal Distribution of Indian
Nummulites, Rec. Geol. Surv. India, vol. xxxiv. (1906),
p. 79. (2) Distribution of Orthophragmina and Lepido-
cyclina, ibid., vol. xxxv. (1907), p. 62.

X

THE VEGETABLE KINGDOM

PLANTS are divided by botanists into four grades, corre-
sponding broadly to an advance from an entirely aquatic
to an entirely terrestrial life. From what we have
learned from fossil animals we may expect that such an
advance would not take place along one line but along
several parallel lines, and that there would be occasional
reversions to an aquatic life on the part of plants whose
ancestors had become adapted to the land. Both these
expectations are justified.

" i. Thallophyta. — This, the lowest grade, corresponds
in a general way to the Protozoa among animals, its
members being composed of cells only, though they may
form aggregates of much greater size and more dif-
ferentiation than do the Protozoa. An alternation of
sexual and asexual generations (as in the Foraminifera
and other Protozoa) exists, and is important for its
bearing on the life-history of the higher plants. The
Thallophyta include the Algae (the great majority of all
aquatic plants) and the Fungi.

2. Bryophyta. — In this grade (mosses and liverworts)
the alternation of generations is much more regular
than in the Thallophyta, the sexual generation (garnet:

33i

332 PAL/EONTOLOGY

tophyte] being the larger and more important plant, the
asexual (sporophyte) being short-lived and always attached
to the gametophyte : hence it is commonly taken for a
mere fruit of the latter. The male and female repro-
ductive cells can only function in a liquid environment,
hence the gametophyte can only live in damp situations ;
but the asexual spores produced by the sporophyte are
adapted to resist drought and to be disseminated by the
wind. The plant-body of the Bryophyta shows greater
differentiation in its cells than the Algae, and the
beginning of formation of vessels, but it never attains a
truly woody structure and is hence almost unknown in
the fossil state.

3. Pteridophyta. — In this grade (Ferns, etc.) there is
a reversal of the relative importance of sporophyte and
gametophyte. The former is now the larger and longer-
lived, it is well differentiated into root, stem, and leaf, and
has woody and other tissues capable of preservation in the
rocks. It can live in dry soils, but must not be far from
water. The spores are borne in spore-cases (sporangia)
usually on the ordinary leaves, sometimes on special
leaves (sporophylls\ and are wind-scattered, germinating
into gametophytes in damp places.

The gametophyte is very small, delicate, as simple in
structure as any of the Bryophyta, and short-lived. Some-
times the sporangia and spores are of two kinds, large
megaspores from which female gametophytes grow, and
small microspores which produce males.

4- Spermaphyta (Seed-Bearing Plants).— In these,
the highest plants, the megaspore germinates, forms a

THE VEGETABLE KINGDOM 333

female gametophyte, which is fertilized and produces an
embryo of a new sporophyte — all without ever being shed
from the megasporangium, which, when all these develop-
ments have taken place within it, is called iheseed. The
microspores are called .pollen, and germinate on the
megasporangium. Thus the dependence of the gameto-
phyte on an aquatic (or even a damp) environment is
entirely avoided, and adaptation to the driest conditions
becomes possible.

Algae. — Most of the Algae leave no trace in the
rocks except indefinite carbonaceous remains. There
are two groups which form exceptions to this rule.

1. The Diatomacece (diatoms), microscopic forms with
a siliceous skeleton, with a very fine network-structure,
resembling that of Radiolaria but much finer. These
occur in immense numbers in fresh and salt waters, and
their remains accumulate to form masses of diatomaceous
ooze, in the cold Arctic and Antarctic seas especially.
Similar deposits occur in Tertiary and Cretaceous rocks,
and are worked for polishing powder and as the inactive
basis of dynamite. The deposits of Bilin (Bohemia)
and Richmond (Virginia) are examples. Sometimes the
deposit becomes a chert. Below the Cretaceous rocks
they are practically unknown.

2. Among the red and green seaweeds there are a
number of families in which calcium carbonate is
deposited in the cell-walls. These forms are collectively
known as " calcareous algae," and some of them have
been shown to play a large part in the building of
modern coral-reefs, and they were no less important as

334 PALEONTOLOGY

rock-builders in the past. Their remains have often
been misunderstood and attributed to Foraminifera,
Sponges or Hydrozoa. The calcareous green seaweeds
(Chlorophyceae) form in general branching tubular
structures, which often break up into segments of
characteristic appearance. Such are Gyropovella (Perm.-
Trias.) and Diplopora (Trias.), which play a part in the
building of the Alpine dolomites, and Ovulites (Eocene),
of which the segments, common in the calcaive grassier of
Paris, are egg-shaped with a central tubular cavity.

The calcareous red seaweeds, on the other hand, are
more often massive or encrusting. Important are
Solenopora (Camb.-Jur., Fig. 100, a), which contributes to
the substance of many British Palaeozoic Limestones, and
Lithotliamnion (Cret.-Rec.), a rock-former in the Miocene
of the Vienna basin.

An interesting group, which attains a higher grade
than the true Algae, though on different lines from
other plants, is the Characea, of freshwater habitat and
secreting a calcareous skeleton. It is best known by its
spirally-marked fruits, found in the Purbeck and Oligocene
beds of the South of England (Fig. 100, b).

Pteridophyta (Vascular Cryptogams). — Three distinct
lines of descent may be recognized, in all three of which
two sub-grades are passed through (a lower homosporous,
with only one kind of spore and sporangium, and a
higher heterosporous, with mega and micro-sporangia and
spores), while two of them evolve independently into
the higher seed-bearing grade. These three lines are
the Lycopodiales (represented to-day by the creeping,

THE VEGETABLE KINGDOM 335

moss-like Lycopodium and Selaginella) the Equisetales (by
the marsh-loving " horse-tails ") and the Filicales (ferns).
The two first, at least, existed in the earliest known land
flora (Devonian) and attained their highest development
in the newer Palaeozoic, and alongside them are found
both undoubted ferns and another group, Sphenophyllales,
which appears to represent the common stock from which
they sprang. The extensive working of coal in the
Carboniferous rocks has enabled the flora to be collected
very extensively and studied very thoroughly, and some
explanation is necessary of the conditions of preservation
of these fossils.

The remains of land-plants are of two kinds — impressions
of leaves, bark, etc., which show in a carbonaceous film
the outward form and surface- markings, often very
perfectly, but preserve no internal structure ; and petrifac-
tions, either in silica or calcite, which preserve internal
structure often as perfectly as in a recent plant kept in
methylated spirit, but may often fail to show outward
form. Further, the various parts of a complete plant —
root, stem, leaves, sporophylls, seeds — are rarely found
in organic continuity, and even such a single part as the
stem may present very different appearances according
as the outer cortex is preserved or not, etc. It is evident
therefore that the palaeobotanist has to reconstruct his
plants out of very scattered and fragmentary evidence.
Great difficulties of nomenclature arise : leaf impressions
receive one set of names, stems another, roots another,
and even when leaf, stem, and root have been pieced
together it may be found that leaves whose forms differ

336

PAL, EON TO LOGY

FIG. 100. — (For description seep. 337.)

THE VEGETABLE KINGDOM 33?

so much that they have been put into distinct geriera
belong to stems which cannot be distinguished, and
again stems which seem justifiably to be divided into
two or more genera have indistinguishable roots. A
triplicate nomenclature at least is therefore inevitable,
and palaeobotanists speak of " leaf-genera," etc. In the
case of the most fully investigated Coal- Measure plant, it
is known that at least five generic and five trivial names
had been given to its separate parts.

Sphenophyllales. — Only one genus is known, Spheno-
phyllum (Dev. ? Carb.— L. Perm.), which seems to have
been a climbing plant, very much like the modern bed-
straws in appearance. The leaves are in whorls round
the stem, and are wedge-shaped, narrow at the attachment,
widening out and ending abruptly at their broadest.

Equisetales. — This group attained its highest develop-
ment in the Carboniferous period, when it was represented

FIG. 100. — FOSSIL PLANTS.

a, Solenopora compacta (Billings), Ordovician. a, Vertical section (X25),
showing part of three concentric layers of cells. a', Horizontal
section. (X5O.) (After Brown.) b, Chara lyelli Forbes, Oligocene.
Fruit. (Xio. ) (After Forbes.) c, Sphenophyllum schlotheimi Brong-
niart, Coal Measures, c, Stem with whorls of leaves. (X£.)
c' , single leaf. (Natural size.) d, Lepidodendron sternbergi Brong-
niart, Coal Measures. (Xj.) d', Leaf-traces of L. gracile Lindley.
(X^.) e, Sigillaria tessellata Brongniart, Coal Measures, e, Part of
trunk (X§.) e', Leaf- traces. (X±.) /, Calamites mougeoti Brong-
niart, Coal Measures. Trunk. (X,.) /', C. decoratus Brongniart,
Coal Measures. Bark. (Xj.) g, Neuropteris heterophy lla Brongniart,
Coal Measures. Part of leaf. ( X J). g*, Venation of N. blissi
Lesquereux. (X§.) (After Kidston.) h, Alethopteris lonchitica
(Schlotheim), CoallMeasures. Leaf. (x£.) ti, Leaflet. (Natural size.)
To show venation, z, Glossopterisbrowniana Brongniart, Permo-Carbo-
niferous, New South Wales. Leaf. (Xj.) i', Venation. (Xf.)
(After Seward. ) j, Williamsonia pecten (Phillips), Lower Estuarine
beds (Aalenian), Scarborough. (Xj.) (After Seward.) k, Ginkgo
digitata (Brongniart), Middle Estuarine beds (Bajocian), Scarborough.
(X^.) (After Seward.) (After Brongniart or Lindley, where not
Otherwise stated.)

22

33& PALAEONTOLOGY

by trees which attained a height of fifty feet or more and
a thickness of several feet, the latter attained by a
process of secondary thickening like that of modern
trees. Like the little Equisetum of to-day, they had a
hollow stem, and are most frequently represented by
internal casts (pith-casts), the markings on which are
very like those of the exterior of Equisetum— vertical
flutings, alternating in position at the nodes (where
leaf- whorls arose). Chief genera: Archceocalamites (U.
Dev.-L. Carb.), Catamites (U. Carb., Fig. 100, /, leaves
known as Anrmlaria, Asterophyllites, Calamocladus, repro-
ductive cones as C alamo stachys, etc.), Schizoneura (Perm.-
Trias.), Equisetites (Rhaetic-Wealden).

Lycopodiales. — The post-Carboniferous members of
this group are all herbaceous plants, of little geological
importance, but in the Devonian and Carboniferous
periods it was also represented by forest-trees of great
size and with secondary thickening. Genera : Lepidoden-
dron (Dev.-Carb., Fig. 100, d), Sigillaria (U. Carb., Fig.
100, e). These two genera are readily distinguished by
the leaf-scar pattern on the bark — the former having
large rhomboidal scars covering the whole surface, the
latter having small scars of varying shape in vertical
rows: in both, the full pattern is only shown on the
exterior of the bark, its internal surface or the decorti-
cated stem showing a ghost of the same pattern. The
roots of both are alike, branching dichotomously (Stig-
maria)', the reproductive cones of Lepidodendron are known
as Lepidostrobus, those of Sigillaria as Sigillariostrobiis.
In some species of Lepidodendron the megasporangium

THE VEGETABLE KINGDOM 339

seems to have become a true seed, the microspores
germinating on it and fertilization taking place inside it.

Filicales. — The modern ferns are divided into three
groups, (i) Eufilicinae, which includes all the familiar
ferns : they are described as homosporous (producing one
kind of spores only) and leptosporangiate (the sporangia
developing from a single cell) ; (2) Hydropteridae, a small
group of water-ferns, which are leptosporangiate but
heterosporous (with mega- and micro - sporangia) ;
(3) Marattiales, a very small group which are euspor-
angiate (sporangia developed from a group of cells) and
homosporous.

Fossil Eufilicinae, related to the recent Osmunda, are
known from Upper Permian upwards, and represen-
tatives of other families from the Upper Triassic
upwards. In the various estuarine beds of the Jurassic
of Yorkshire, for instance, there are found fern-leaves
closely similar to those of ferns now found only in the
Malayan region. Hydropteridae are known from the
Eocene upwards, but from earlier times the only records
are very doubtful.

The Marattiales, however, are very well represented,
from the Carboniferous onwards. Some at least of the
species placed in the leaf-genus Pecopteris (Carb.-Jur.)
seem to belong to the tree-fern stem Psaronius (U. Carb.-
L. Perm.). Other forms are known from Mesozoic and
Cainozoic rocks.

But the great majority of what have been called ferns
in the Carboniferous flora have proved to belong to a
higher grade. They may be related to the Marattiales

340 PALEONTOLOGY

as the Hydropteridae are to the Eufilicinae, but they had
advanced much beyond merely being heterosporous, and
form the first division of the Spermaphyta.

SPERMAPHYTA.— The seed plants again fall into
two sub-grades, according to whether the seed is borne
on the surface of a sporophyll (Gymnosperms) or in a
closed ovary formed by the folding over of one or more
sporophylls (Angiosperms). The Gymnosperms include
the following five groups.

1. Pteridospermeae. — This extinct group is inter-
mediate between the eusporangiate ferns (Marattiales)
and the cycads. It is confined to the Upper Palaeozoic,
and best known from the fern-like leaves of the Coal
Measures. Only of recent years, by the piecing together
of evidence from the microscopical structure of petrifac-
tions with that from leaf-impressions, has it been possible
to discover its true position. Common leaf-genera are
Alethopteris and Neuropteris (Fig. 100, g, h), corresponding
to the stem genus Medullosa and the seed Trigonocarpmu ;
Sphenopteris, to the stem Lyginodendron and the seed
Lagenostoma. Other leaves probably belonging here are
Glossopteris (Permo-Carb.-Rhaetic, Fig. 100, *) with its
rhizome Vertebraria, and the allied Gangamopteris (Permo-
Carb.)

2. Cycadophyta. — The modern cycads are a group of
plants with palm-like foliage, but much more primitive
in their sporophylls than the palms (which are angio-
sperms). They are confined to tropical regions, but
extend all round the world. Their Mesozoic allies had

THE VEGETABLE KINGDOM 341

no such restriction in latitude, being common in England
for instance, and some of them attained a far higher grade
of floral structure, which makes it probable that from
them sprang the Angiosperms (typical flowering plants
with seeds enclosed in a fruit). They are doubtfully
represented in the Coal Measures, but undoubtedly from
the Rhaetic, Jurassic, and Lower Cretaceous. The
Jurassic period has been termed the "age of cycads."
Their remains are abundant in the Jurassic estuarine
beds of the North of England, in some of the Purbeck
beds, where they occur in the position of growth in the
" fossil forest " of Dorset cliffs, and in the Wealden beds.
But the most famous locality for them is the Black Hills
of Dakota, where wonderfully preserved specimens of
Upper Jurassic age have enabled Dr. Wieland to deter-
mine the full details of the floral organs. Leaf-genera :
Pterophyllum ( Trias. ),Cycadeoidea, Williamsoma(Fig. 100,7),
Ptilophyllnm (all Jur.), Otozamites (Trias.-L. Cret), Nils-
sonia (Jur.-L. Cret.), Zamites (L. Cret.)

3. Cordaitales. — The genus Cordaites (Dev.- Rhaetic)
was a forest tree resembling in habit the Australian
Kauri pine, to which it may have been not very distantly
related. Its leaves were long, narrow, thick and parallel
veined ; its seeds (Cardiocarpus) were borne on a cone.

4. Ginkgoales. — The maiden-hair tree, Ginkgo, with
its bifid wedge-shaped leaves (Fig. 100, k), is now confined
to parts of China and Japan ; but closely allied forms had
a world-wide distribution in Jurassic times, ranging from
Greenland to Tasmania. Its time-range in Europe is
Rhaetic to Eocene.

342 PALEONTOLOGY

5. Coniferales. — These include the familiar firs and
pines, the yew and monkey-puzzle (A raucana). This last
is now native only in South America and Australasia, but it
also had an almost world-wide distribution in the Jurassic
and Lower Cretaceous. Allied forms are Walchia (Perm.,
rare in U. Carb.) and Voltzia (Perm.-Trias.). Coniferous
wood is a common object in many Jurassic and Cretaceous
rocks, both marine and freshwater.

Angiospermae. — A few rare remains of angiosperms
occur in Lower Cretaceous strata, but at the beginning
of the Upper Cretaceous period there was such a sudden
appearance of an extensive angiosperm flora in Europe,
that it is obvious that such a cryptogenetic flora must
have migrated from some other part of the world, as yet
unknown. Its fossil remains are found, not in England,
which was then covered by a fairly deep sea, but in
regions like Saxony, which were close to a shore-line
and where sandy deposits occur in place of chalk. From
this time onwards, the angiosperms are the dominant
members of the flora everywhere. Their natural orders
and genera are far too numerous to be mentioned in
detail here.

The possible existence of marine life provinces in
past times has been referred to in Chapter V. Fossil
plants have provided the most unquestionable example
of terrestrial life provinces. The flora of Devonian and
Lower Carboniferous times is practically the same all
over the world, but in the Upper Carboniferous there
were two sharply contrasted floras — a northern flora,
including Lepidodendron, Sigillana, Calamites, .and many

THE VEGETABLE KINGDOM 343

pteridosperms, and a southern flora, without any large
trees, but chiefly characterized by Glossopteris and Ganga-
mopteris. The former is found over most of the northern
hemisphere, both in the Old and New Worlds, the latter
is principally found in the southern hemisphere and
India. So entirely different are the two, that for a long
time many geologists would not admit them to be con-
temporaneous, but held the Glossopteris flora as Mesozoic,
it having in fact nearer affinities to the European Mesozoic
than to the European Carboniferous flora. The discovery
of marine strata having an unmistakable Upper Carboni-
ferous or Lower Permian fauna, alternating with beds
containing the Glossopteris flora, at length gave conclusive
proof of the Permo-Carboniferous age of the latter. It
appeared, therefore, that there were two continental
masses, a northern on which the descendants of the
Lower Carboniferous plants continued their evolution,
and a southern on which that flora had been exterminated .
(possibly by glacial conditions of which there is good
evidence), and a new and much less rich flora developed.
Later discoveries have shown that there was some
geographical overlap of the two floras, Sigillaria having
been found in South Africa, and Glossopteris in Russia ;
but the exact nature of this overlap remains to be ex-
plained. By the Rhaetic epoch a uniform flora had once
again established itself over the land-surfaces of the
world.

344 PALEONTOLOGY

Short Bibliography.

ARBER, E. A. NEWELL. — (i) The Glossopteris Flora,
Brit. Mus. Cat. (1905). (2) Several papers on Coal
Measure Plants in Phil. Trans. Royal Soc. (1904-16). (3)
The Natural History of Coal (Cambridge University
Press, 1911).

GARWOOD, E. J. — Calcareous Algae (Pres. Address
Brit. Assoc., 1913), Geol. Mag. (5), vol. x. (1913).

HICKLING, G. — Contribution to the Micro- Petrology of
Coal, Trans. Inst. Mining Engineers^ vol. liii. (1917).

KIDSTON, R. — Many papers on Coal Measure Plants
in Trans. Roy. Soc., Edinburgh.

SCOTT, D. H. — (i) Studies in Fossil Botany (2nd
edn., 1909). (Microscopic Structure of Coal Measure
Plants.) (2) Presidential Addresses, Royal Micr. Soc.
(1905-07).

SEWARD, A. C. — (i) Fossil Plants, vols. i.-iv. Cam-
bridge Biological Series (1898-1919). (2) The Jurassic
Flora, Brit. Mus. Cat., vol. i. (1900), ii. (1904).

STOPES, M. C. — (i) Ancient Plants (1910). (2) Studies
in the Composition of Coal, Proc. Roy. Soc. B., vol. xc.

WIELAND, G. R. — American Fossil Cycads, Carnegie
Institution of Washington (1906).

XI

THE COLLECTION AND PRESERVATION OF
FOSSILS

REFERENCE has incidentally been made in previous
chapters to some of the conditions of preservation of
fossils. They may here be summarized :

1. Very rarely is an animal or plant body preserved
whole : the case of insects in amber — a fossil resin which
enveloped them as it trickled down the bark, preserved
them by its antiseptic character, and became hardened
by loss of its volatile constituents — is perhaps the only
case. The frozen mammoths of Siberia are not true
fossils, as they are not entombed in sediments.

2. While the soft parts decay, the hard parts — mainly
of mineral substance, but sometimes organic — may be
preserved without any chemical change. This is a very
common case among animals, if we interpret the term
"chemical change" rather broadly. Some amount of
chemical change is almost inevitable — the filling-up of
minute cavities with secondary mineral matter, as in
echinoderm skeletons and vertebrate bones, is rarely
escaped. Among plants, where the hard parts are
organic, some carbonization nearly always takes place ;
but in the paper-coal of Tula, in Russia, the cuticles of
the plants are said to be absolutely unchanged.

3. The hard parts may he preserved, but with more or
less complete chemical change. If the original skeleton
was of calcite this may be replaced by silica, dolomite,

345

346 PALEONTOLOGY .

chalybite, limonite, calcium phosphate, pyrite or selenite.
Aragonite may be replaced by any of the same, or may
undergo a paramorphic change into calcite. The extent
to which these changes may alter the minute structure of
the skeleton varies very much : when any recrystalliza-
tion takes place the minute structure is usually destroyed,
wholly or in part. Amorphous silica (opal) may preserve
the finest details, but if it passes into chalcedony it will
often produce a structure due to imperfect crystallization
in rings (beekite structure), which will not only destroy
minute structure but even obscure the outward form.
So, too, with the formation of selenite (gypsum), few
fossils but belemnites can survive it in recognizable
form ; but this chemical change differs from most others,
as it is probably quite a recent result of weathering near
the surface, while most others took place soon after the
entombment of the fossil.

Originally siliceous skeletons may be altered to calcite,
pyrite or marcasite, the last in turn passing into limonite
under conditions of weathering. Sponges in the chalk
may sometimes be found, partly silica, partly limonite.

Hard organic tissues generally undergo some degree of
carbonization, but they may be replaced by pyrite (as
with graptolites), or by calcite or silica (as in plant
petrifactions).

Phosphatic skeletons are the most stable of all, and
beyond having their minute cavities filled with secondary
phospate they undergo no change. Very rarely they may
be changed to vivianite (a blue iron phosphate).

4. An organism may be represented by casts of its
skeleton — external casts or moulds, and (if a hollow
skeleton) internal casts. External casts are composed
of the matrix of the rock, that is to say the sediment in
which the organism was buried : in hard enough rocks
they are always produced when a fossil is extracted, but

COLLECTION AND PRESERVATION OF FOSSILS 347

they may be found naturally produced through the
removal of the fossil in solution. Infillings of the
interior may also consist of the matrix, when that had
free access to the interior ; when it had no access, they
consist of some material deposited from solution, or
partly the one and partly the other, when the sediment
could only penetrate into part of the interior. Very
rarely does a hollow skeleton remain- really empty,
though infilling may often be incomplete. In limestones,
the material deposited from solution will be calcite, or
chalcedony (as in the chalk-flints) ; in clays it is usually
pyrite, or marcasite (the two forms of FeS2). In cases
where deposit on the sea-floor was extremely slow,
calcium phosphate is a usual infilling, or for minute
cavities glauconite : these two substances being deposited
before burial is complete.

In any of these cases it is only by subsequent removal
of the shell that the infilling appears as an actual
internal cast. Such removal may be by solution on the
sea-bottom or within the rock, or it may be artificially
produced, either accidentally during extraction, or de-
liberately for the purpose of investigating internal
structure. When naturally produced, the external and
internal casts together are often described as "hollow
casts": by pouring melted wax or gutta-percha, etc.,
into the cavity, a mould of the fossil can be produced.

5. There are other traces of living organisms in rocks
that may rank as fossils, though even less definite than
casts. Such are footprints and other impressions of
moving animals in sand or mud (preserved by being
covered by a lamina of a different kind of sediment),
burrows in sand or mud (preserved by being filled with
slightly different material) and borings in hard rocks or
in fossils. Among boring organisms are minute algae,
sponges (Cliona), some echinoids, many lamellibranchs

348 PALEONTOLOGY

(Lithodomus, Pholas, Teredo). Sometimes when a shell is
bored into, the material filling the boring may resist
subsequent solution better than the shell itself, and a
cast of the boring may result.

All these various ways of preservation must be borne
in mind by the intelligent collector when at work on
cliffs and quarries. While the spoil-heap of a quarry
will often provide the choicest specimens, he should
never neglect to search for fossils in situ in the undis-
turbed rock. Only by so doing can he know whether all
the fossils belong to one fauna, or whether there are
several, and through what thickness a single fauna per-
sists ; only so can he see whether lamellibranchs, for
instance, are in the vertical position of living burrowers
or in the horizontal position of drifted shells ; only so can
he distinguish with certainty the presence of derived
fossils, whose original home was in some older bed and
whose worn and rounded surface distinguishes them
from the contemporaneous fossils with which they may
be mixed.

The collector would be well-advised not to attempt to
clean or select his fossils on the spot where he obtains
them, except when their weight is serious. Many a good
specimen has been spoiled by the attempt to clean it on
the spot. Even broken specimens may often prove to
show special features not seen in perfect examples. The
writer once picked up seventy-three specimens of
Chonetes laguessiana from the Carboniferous shales of
Fourstones (Northumberland). They all looked alike
at the time, but on cleaning and sorting them at leisure
afterwards, fifty-eight were found to be complete speci-
mens but with the dorsal valve crushed in ; fourteen were
ventral valves, and one was a dorsal valve showing both
external and internal features. That one dorsal valve
would probably have been lost had any attempt to clean
or select been made on the spot.

COLLECTION AND PRESERVATION OF FOSSILS 34$

Care must be taken to pack up specimens so that they
do not get broken in transit home : heavy fossils must
not be placed on top of fragile small ones, for instance.
They must be so labelled that there will be no mistake as
to the exposure and the particular bed from which they
came. At the time it may seem quite safe to trust to
memory, but a week afterwards an unlabelled specimen
will prove a serious worry to the conscientious collector.

The extracting and cleaning of fossils — the removal of
fossil from matrix or matrix from fossil — is an art in
itself. In extracting large specimens from hard rocks,
such as tough limestones, the latter must be subjected to
great stress — the blow of a sledge-hammer may make a
fossil jump out quite clean. A screw-movement stone-
breaking machine may be effectively used in the same
way. At other times when the fossils are fragile and
firmly cemented to the matrix, their violent extraction
may be almost hopeless, but heating the rock and cooling
it by plunging it into cold water may have the required
effect. In the extraction of small fossils, such as fora-
minifera and ostracods, from sands and friable rocks, the
rock is first broken up and then passed through a series of
sieves : the great majority of fossils will be found between
a mesh of 32 to the inch and one of 64 to the inch. In
the case of stiff clays, another method must be used :
the clays are broken up into small pieces, about the size
of a pea or larger, and heated on a metal plate over a
flame until thoroughly dry, when they are dropped while
still hot into cold water in a circular dish. The clay
quickly disintegrates into mud, and by keeping up a-
swirling of the water by constant movement of the dish,
decanting the muddy water and adding fresh over and
over again, all the argillaceous material is at length
removed and a concentrate left, consisting of sand-grains
and small fossils. This is dried and then spread out, a

350 PAL/EONTOLOGY

little at a time on black paper, examined with a lens and
fossils picked out with a moistened sable brush, or in the
case of heavy ones with a fine forceps. Siftings of friable
rocks are picked over in the same way.

Chalk may be scrubbed under water with a stiff brush,
and a concentrate obtained by the method just described:
the chalk from the interior of large fossils (such as
echinoids) is often the most fruitful, because the small
shells have been more protected from destructive move-
ments in the ooze on the sea-bottom.

The methods of removing of matrix from a fossil
depend greatly on their respective toughness. When
the fossils are strong and the matrix friable, washing
with or without the help of a brush (a tooth-brush is very
convenient) may be enough ; but some fossils (many
lamellibranchs) will break into pieces if wetted, and must
be cleaned dry with the help of some tool. A shoe-
maker's awl mounted in a handle is very useful for
tough matrix ; in other cases a mounted needle, pin, or
steel pen, or even a common pocket-knife may be service-
able. The golden rule about all such tools is that they
should never be used to scrape the matrix away from the
fossil (with a possible exception for silicified and pyritized
fossils, where the fossil is harder than the steel of the
tool). The aim should always be to put such a strain on
the matrix by means of a tool, that it breaks away from
the fossil without the tool touching the latter. The
application of this rule to particular cases must be very
varied.

As an example, take the case of a graptolite, part of
which is exposed on the lamination surface of a shale,
partly buried under other laminae. The natural first
idea is to push a knife-blade between the laminae and
separate them, but the probable result of this will be that
the_ laminae break away anywhere except where the

-COLLECTION AND PRESERVATION OF FOSSILS 351

graptolite lies, and in the end the hidden part may be
broken before it is exposed. If, however, we press the
knife-blade vertically downwards along a line parallel to
the hidden part but a quarter-inch or more away, the
strain on the concealing laminae of shale will very likely
cause them to split off just where wanted.

Brushes, the bristles though stiff being softer than
most fossils, can be applied directly to the surface of
fossils. The dental drill used by dentists for the filing
of teeth can be adapted to the cleaning of fossils: it
enables a rapidly rotating brush to be applied to any
point on the surface of a fossil, with much more con-
centrated and rapid effect than that of a hand-brush.
Wire brushes can even be used to clean pyritic fossils
embedded in hard slates.

Where mechanical means fail, chemical methods may
be used. The simplest of these is ordinary weathering,
which in many cases disintegrates matrix without injur-
ing fossils (though in other cases the reverse is the case).
Very weak acids may effect the same result more rapidly.
In general acids must be resorted to with great care,
since most fossils are calcareous and at once attacked by
them. Where it is desired to expose more fully a fossil
partly embedded in tough limestone, the exposed surface
may be varnished to protect it and acid poured on :
careful watch must be kept so that as soon as any more
is exposed it may be washed and varnished.

A generally safer reagent is caustic potash. This is
applied dry, small pieces being scattered over the surface
of matrix which it is desired to remove : as the potash
deliquesces it penetrates into the matrix and disintegrates
it, while it does no harm to the fossil if it be an echino-
derm or a mollusc. (It must not be used, however, for
brachiopods, their laminated shell being easily penetrated
and flaking away in consequence.) The fossil is then

3$2 PALEONTOLOGY

thoroughly washed and scrubbed, and it is well to preserve
the washings and evaporate them to a more concentrated
solution of potash, to be used for another purpose
mentioned below.

Many fossils, though stable enough in their long
home, are physically or chemically unstable as soon as
extracted, and need special treatment to save them from
disintegration or the risk of it. If they are simply fragile
they must be impregnated with some binding material.
The most usual one is gelatine, applied as a hot, strong
solution with a soft brush : it penetrates the shell and as
it hardens it binds the parts together and prevents decay.
It has the disadvantage of requiring very rapid treatment
lest it harden before penetrating. Another material
is a solution of collodion in amyl acetate : this has the
advantage that it hardens much more slowly by evapora-
tion. It can be brushed over the shell at intervals of a
day or more until no more is absorbed, or small specimens
can be dropped into the solution and left there indefinitely.
Care must be taken not to inhale the amyl acetate.

The chief case of chemical disintegration is that of
fossils or casts in marcasite. This is an unstable sulphide
of iron which absorbs moisture and oxygen, and gives a
crystalline efflorescence of ferrous sulphate. As soon as
such an efflorescence is noticed, it must be brushed away
and the fossil soaked in hot, but not boiling, potash
solution to neutralize any free sulphuric acid. It is then
washed thoroughly in distilled water to remove the potash,
and thoroughly dried on a hot plate, after which it ^s
painted all over with shellac varnish (or if not too large
soaked in the varnish). In this way further access of
oxygen and moisture is prevented.

When a fossil is represented only by a cavity in a rock,
it becomes necessary to take a cast of it. For this
purpose many substances may be used — paraffin wax,

COLLECTION AND PRESERVATION OF FOSSILS 353

sealing wax, gutta-percha and the various compositions
used by dentists for moulding the teeth and palate.

The various methods so far described give some idea
of the most frequent and general forms of palaeon-
tological technique. Other special methods can only be
mentioned, such as the delicate methods employed by
Norman Glass and others for exposing the brachial
spirals of brachiopods, the methods of exposing, tracing,
and reproducing the suture -lines of ammonoids (see
Frontispiece), the making of serial sections of solid fossils
at regular intervals and constructing from them enlarged
models of the internal structure. The accurate measure-
ment of fossils by means of sliding callipers is a very
necessary process in the identification of species and the
tracing of gradual evolution.

XII
THE RULES OF NOMENCLATURE

THE subject of the nomenclature, or naming of fossils, is
one productive of much irritation and of many sarcastic
utterances among geologists, to whom the time spent by
many palaeontologists on the accurate determination of
the names of fossils appears as sheer waste. The
attitude of mind of many critics of these palaeontologists
is expressed by a rather hackneyed quotation from
Shakespeare, of which it is sufficient to say that if Juliet
had ordered a rose from her gardener by the name of
garlic, the rose would not have smelled as sweet as she
desired. The whole purpose of giving names to things
is that persons may be able to speak to one another
about them without danger of misunderstanding.

In proportion as we wish to discriminate more and
more minutely between things, we require more and
more exact names for them. There are purposes for
which roses and lilies and buttercups may be spoken of
without discrimination as "flowers"; there are others
for which they require distinct names ; there are yet
others for which the need is felt of finer discrimination,
by means of qualifying adjectives as " red rose " and
" white rose " ; while a modern gardener requires
separate names for a great number of different kinds in
each of these categories.

The naming of fossils is only a part of the naming of
animals and plants in general. Leaving plants aside

354

THE RULES OF NOMENCLATURE 355

for the moment, the Animal Kingdom is divided into
a number of phyla, each of which is divided into classes,
these again being divided into orders, these into families,
these into genera, and these into species. To define a
species is difficult enough among living animals, where
the test of inter-breeding is available ; among fossils
there is only one really scientific way of delimiting
species, and that is by determining whether variation is
continuous or discontinuous. To do this a large number
of individuals from the same bed must be taken, note
made of those measurable characters in which they show
variation, and a series of accurate measurements taken.
The results for each character are plotted as a graph, the
horizontal coordinates being the measurements and the
vertical coordinates the number of individuals giving
each particular measurement. If the plotted curve is a
simple one with a single maximum, that is evidence for
treating all the individuals as one species ; but if it
shows two or more maxima with very decided minima
between, then there is ground for regarding them as two
or more species. Confirmation must be sought in the
curves for the other measurable characters.*

This method is only practicable when large numbers
of individuals from the same bed are available, and
when their variations are expressible in simple numerical
form. In the absence of such conditions no better
definition of a species than this can be suggested for the
purposes of palaeontology : A species is a collection of
individuals so nearly alike that they may conveniently be
denoted by the same. name. This definition leaves the
decision as to whether there is sufficient likeness to the
judgment of every palaeontologist, and judgments will

* For a good illustration of this method see Hickling, G.,
"Variation of Planorbis multiformis Bronn," Mem. Proc. Manchester
Lit. Phil. Soc., vol. Ivii., pt. 3 (1913).

356 PALEONTOLOGY

inevitably differ : as such differences constantly occur,
this definition recognizes the facts of the case.

A genus may be defined either as a collection of species
having certain features in common (the morphological defini-
tion) or as a collection of species believed to be derived from
the same immediate ancestral stock (the genetic definition).
The difference between these definitions has been
illustrated under the Graptolites.

As a rule, each of the categories named above consists
of several of the next category below, but it may consist
of only one — e.g., a genus may include only one species,
a family may include only one genus.

If the categories named are not enough for satisfactory
classification in any particular case, intermediate grades
may be intercalated, denoted by the prefix sub- ; thus an
order may be divided into several sub-orders, each con-
taining several families. Still further grades may some-
times be necessary : they are denoted by the prefix super- :
thus a sub-order may be divided into super-families, and
these into families. The term variety is generally used
in place of sub-species.

The system of nomenclature for species now universally
adopted is that of the Swedish naturalist Linne (latinized
as Linnaeus) and consists in denoting every species by a
double name: hence it is often called the binominal
system. Thus "Dalntanites caudatus" is the specific
name of a certain species of Trilobite, and is made up
of the generic name " Dalmanites " (the name of the
genus to which this species belongs), and the trivial
name " caudatus " which distinguishes this species from
others of the same genus.

A generic name is always a single word, either a Latin
noun or a word treated as such. In the early days of the
Linnaean nomenclature Latin or latinized names of
animals were plentifully at hand, but with the increase

THE RULES OF NOMENCLATURE 357

in the number of genera new kinds of names had to be
sought. Names of mythological characters, such as Venus
and Action, were appropriated, but have long been
exhausted. Then came more or less descriptive com-
pound words of Latin or Greek derivation, such as
Micraster (little star), and Macrodon (big tooth) : Many
genera were named in compliment to individuals, as
Muvchisonia and Lonsdaleia, or from the locality where
they were first discovered, as Amberleya, Bohemilla. As
the number of generic names increased enormously there
came a natural tendency to help the memory by com-
pounding similar names for genera in the same class ;
thus the great majority of crinoid genera have names
ending in -crinus, those of many corals end in -phyllum or
phyllia, of many starfish in -aster, of cephalopods in -ceras,
of brachiopods in -thy vis. But there is no copyright in
such endings and they must not be trusted as a necessary
indication of affinities. Thus Cvyptocvinus is not a crinoid
but a cystid ; Holocystis, not a cystid but a coral ; Campto-
ceras and Platyceras are gastropods. Most names ending
in ~obolus denote brachiopods, but Trochobolus is a sponge.
The attempt to combine the name of a man or a place
with one of these terminations has led to such uncouth
compounds as A gassizocrinus and Quenstedtoceras.

It was at one time a rule that the generic name of any
fossil should bear the termination -ites. Thus a fossil
Nautilus was called Nautilites, not because it was thought
to belong to a different genus from the recent Nautilus ,
but merely to emphasize its fossil state. This plan has
long been abandoned, but it has led to some confusion
in the case of genera entirely extinct. Many modern
authors have removed the termination in such cases, thus
Echinosph&rites has been altered to Echinosphcera. Un-
fortunately the same termination has been used as a
means of distinction between different genera — Dalmanites

358 PALAEONTOLOGY

was so named, for instance, because its original form
Dalmania was preoccupied. Hence no rule of universal
application can be laid down for such cases, unless it be
the strict rule of priority.

If a sub-generic name is needed it is written in paren-
theses after the generic name; thus Lichas (Corydocephalus)
anglicus denotes that the species belongs to the sub-genus
Corydocephalus of the genus Lichas.*

The best trivial name is a simple Latin adjective,
either descriptive of some feature of the species (as in
Dalmanites caudatus, Belemnites minimus) or indicating the
locality where it was first found (as Crania parisiensis,
Lichas hibernicus, or the much less pleasing oystev-
mouthensis and czenstochovensis). The adjective must
always agree in gender with the generic name, and must
be altered if necessary with a change in the latter.
Instead of an adjective a noun may be used " in apposi-
tion " according to Latin usage, when it keeps its own
gender, as Psiloceras planorbis (wrongly altered by some
authors to planorbe as though it were an adjective),
Agnostus rex, Rhynchonella vespertilio. Thirdly, it may be
a noun in the genitive case, usually (but not always) a
personal name, as Lingulella davisi, Asteroceras smithi. The
rules of the International Zoological Congress only allow
a person's name to be made into a trivial name by the
addition of the letter *, thus not only insisting on some
dreadful barbarisms,! but tending to cause great con-

* Square brackets are sometimes used for a different purpose —
the name which they enclose is not a part of the correct specific
name, but is a reminder that a change of name has taken place —
thus, Basilicus [Asaphus] tyrannus recalls the fact that Basilicus
tyrannus has long been known as Asaphus tyrannus, and is an assist-
ance to those who might fail to recognize a familiar species under
a new generic name.

f For instance, Zaphrentis omaliusi (instead of omalii). But reasons
for making this rule will be found in the Introduction to Bronn's
Index Palaontologicus.

THE RULES OF NOMENCLATURE 359

fusion for the following reason : the early palaeontologists
recognized two ways of dedicating a species to a man,
thus d'Orbigny named two ammonites Ammonites sauzei
and Ammonites sauzeanus. Many modern palaeontologists
insist upon altering all names of the latter pattern into
the former, a proceeding creative of much confusion.
Whatever rule may have to be kept for the future it
best to retain names of old date unaltered in form.

If a sub-species or variety needs to be distinguished,
its varietal name should be written immediately after the
trivial name, as Agnostus pisifonnis obesus ; but many
authors prefer writing Agnostus pisiformis var. obesus.
When a species can be traced, with gradual changes,
through a number of zones, each zonal form is called a
mutation, and may be denoted by the symbol of its zone
(e.g. Pvoductus concinnus Sowerby, mut. D2) or by an
adjective (using mut. instead of var.)

In view of the mistakes that may arise from over-hasty
identifications of a known species at a new locality or
horizon, modern palaeontologists frequently make use of
qualifying or cautionary terms. Their lists frequently
contain such records as Hildoceras cf. Ufrons (Bruguiere)
or Productus aff. productus (Martin). The first means "a
species of Hildocevas very much like bifrons but probably
not exactly identical with it"; and the latter "a species
of Pvoductus which is certainly not identical with P. pro-
d^lctus, but appears to be genetically related to it."

As to the mode of writing or printing specific names,
the following rules should be most strictly adhered to :
(i) A generic name should always bear a capital initial
letter; (2) a trivial name should always bear a small
initial letter* ; (3) when a generic or specific name comes

* To this last rule most continental authors allow an exception
in the case of trivial names derived from a personal name, but the
tendency in England is to make the rule absolute, thus writing
Asaphellus homfrayi in spite of the fact that the trivial name was
given in honour of a Mr. Homfray.

360 PALEONTOLOGY

in the course of an ordinary sentence it should be printed
in different type from the rest of the sentence (normally,
in a sentence of roman type it would be in italics). A
few examples will show the importance of adhering to
these rules. If in a geological work we come across an
allusion to " the Planorbis beds," we should know that
it referred to beds in which there were abundant gastro-
pods belonging to the freshwater genus Planorbis; but
when we read of "the planorbis-beds," we understand
marine beds characterized by the ammonite Psilocevas
planorbis. The Producttis-limestonQ is a well-known
Asiatic formation characterized by the brachiopod genus
Productus ; but an allusion to "the productus-limestone"
would set us puzzling our brains for the genus which
had a species with the trivial name productus. As to the
third rule above, it is only necessary to hunt for the name
of a fossil in some work which has not followed that rule
to realize its importance.

When several species of a genus are referred to in the
same paragraph, it is permissible to use the initial letter
only of the generic name, to save repetition.

If no author had ever, by accident or otherwise, broken
the rules of nomenclature, the generic and trivial names
would suffice to define a species absolutely. It has been
found necessary, however, to avoid confusion, to write
after the name of a species the name of its author — i.e., of
the palaeontologist who first applied that name to it ; and
it is often useful to add the date of his giving the name,
thus: Terebratula intermedia J. Sowerby, 1812,

There is no rule as to the type in which the author's
name should be printed, but it is not desirable that it
should be in the same type as either the specific name or
the rest of the sentence in which it occurs.

The need for further rules arises from the facts that
authors writing at different times have from time to time

THE RULES OF NOMENCLATURE 361

(a) given the same name to different genera, or to different
species of the same genus, either through a mistake in
identification or in ignorance that the name was already
in use ; (b) given different names to the same genus or
species ; (c) divided up an existing genus into several
genera, or (d) united distinct genera into one.

The most general rule in all nomenclature is the law of
priority, according to which, other things being equal,
the earliest name given to any genus or species must
stand : all later names are synonyms, and as such are
rejected. To this rule there are certain necessary quali-
fications :

1. The name must have been published — i.e., printed
and circulated so as to be accessible to the public, and it
must have been accompanied by either a description
(diagnosis) or a figure (or both) which would make it
possible to identify the species to which the name was
applied. In the case of a new genus, a diagnosis is not
absolutely necessary, if the type-species is given.

2. It must have been applied in accordance with the
Linnaean system. This rules out all names published
prior to the year 1758 (date of the tenth edition of the
Systema Nature of Linnaeus) even if they are binominal
in form ; for in pre-L/innaean names that form is accidental,
the earlier method having been to give a generic name
followed by what was really not a trivial name but a
description or diagnosis (though occasionally a single
adjective sufficed).

3. That the same name has not already been applied —
if generic, to some other genus ; if trivial, to soifie other
species of the same genus. In case of such previous
application the name (in its new application) is a homonym.
Thus the name Avalonia was given by Walcott in 1889 to
a genus of trilobites ; the same name was applied by
Seeley in 1898 to a reptile. Hence Avalonia Seeley, 1898,

362 PALAEONTOLOGY

is a homonym of Avalonia Walcott, 1889. Again James
Sowerby in 1820 gave the name Spirifer pinguis to a
Carboniferous fossil ; Zieten in 1838 gave the same name
to a Jurassic fossil. Hence 5. pinguis Zieten, 1838, is a
homonym of S. pinguis J. Sowerby, 1820. When a
homonym has to be quoted, it is advisable to make its
homonymity clear by putting it thus :

Avalonia Seeley, 1898, non Walcott, 1889.

Spirifer pinguis Zieten, 1838, non J. Sowerby, 1820.

Generic homonyms are of course suppressed as soon as
discovered, and a new name substituted — usually a name
only differing slightly from that suppressed. Thus
Emmrich in 1844 gave the name £>almania (in honour of
the Swedish palaeontologist Dalman) to a genus of trilo-
bites which he separated from Phacops ; but Dalmania had
been previously applied to an insect, therefore Barrande
in 1852 substituted the name Dalmanites. Sometimes an
author is less fortunate. Thus Goldfuss in 1839 gave
another trilobite-genus the name Brontes ; finding this
preoccupied he changed it in 1844 to Bronteus, but in the
meantime de Koninck had given the same genus the
name Goldius, and although palaeontologists have been
accustomed to the name Bronteus there can be no doubt
that Goldius has priority over it and should strictly be
used in place of it.

As an example of a specific homonym, we may take
the following case : J. Sowerby, in 1818, named a lamelli-
branch Cardita lirata,a.nd in 1819 named another Lutraria
lirata. In 1826 his son, J. de Carle Sowerby, made a new
genus, Pholadomya, to which he transferred both the above
species, each of them thus becoming Pholadcmya lirata (J.
Sowerby). The 1818 species having the prior claim of
the two to this name, it became a homonym in its appli-
cation to the 1819 species, and consequently a new name
was given to it, and it is now Pholadomya fidicula J. de C.

THE RULES OF NOMENCLATURE 363

Sowerby, of which name Lutraria livata J. Sowerby is a
synonym.*

The fate of all specific homonyms is not quite so clear,
because of the possibility that they may be transferred to
other genera. Thus of the two species named Spirifer
pinguis, Sowerby's is now Bvachythyvis pinguis, and
Zieten's is Spiriferina pinguis : there can therefore be
no danger of confusion between them, and both trivial
names may be kept. It is true that the International
Congress has a very drastic rule — " Once a homonym,
always a synonym," meaning that once a name is
suppressed on account of homonymity it can never be
revived, even on transference to another genus; but
while this applies quite clearly to cases where the homo-
nym has had a new name substituted for it while it was
still in its original genus, it is not clear that it applies to
cases where the homonymity was overlooked until the
change of genus was made, as in the case of Spiriferina
pinguis.

One more exception to the law of priority must be
noted. It has been found impracticable to insist on its
covering both Animal and Vegetable Kingdoms simul-
taneously, so botanists and zoologists have agreed, while
each accepting it within his own kingdom, to let the
nomenclature of the two be quite independent. Thus
Zeilleria is the name both of a Carboniferous pterido-
sperm leaf, and of a Jurassic brachiopod. No practical
confusion is likely to arise between the two.

These are the only valid exceptions to the law of
priority. There are one or two other exceptions that
have been proposed but are now quite rightly rejected.
In the first place a name must not be rejected because it

* It must be clearly understood that the term synonym implies
two names for one thing, while homonym implies the same name for two
things.

364 PALEONTOLOGY

is inappropriate : to allow this would involve us in endless
confusion, since appropriateness is partly a matter of
individual judgment, and if one palaeontologist thinks the
appearance of a certain brachiopod sufficiently suggestive
of a coin to justify him calling it Obolus, while a later
author, failing to see the likeness, is allowed to substitute
another name, who is to settle which name is correct ?
After all, a name is a name, not a description, and in
treating it as such we are following the common-sense
customs of ordinary speech. We have in our houses
kitchen " coppers " made of galvanized iron, drawing-
room fire-" irons " made of brass, and (if still surviving)
bedroom candle-" sticks " made of china; men named
Short or Little are often taller than others named Long,
yet we neither expect them to change their names nor
allow them to do so except with much expense and
trouble ; and no registrar of births would think of object-
ing to such a name as Thomas Thomas on the ground of
tautology, or to Violet Green as a contradiction in terms.
If we remember that the Latin derivation Incus a non
hicendo has become a phrase of general application we
need not be shocked at finding that a crustacean with
an enormous number of legs is called Apus (footless), or
that Agoniatites is not " devoid of an angle" as its name
would imply.

Again, just as the change of manufacture from one
material to another has led to such anomalies as brass
irons, so the transfer of species from one genus to another
has produced curious results. Sometimes a species has
been thought to have only a superficial resemblance to a
genus to which it has afterwards been found to belong,
and the trivial name indicating that resemblance has had
to be kept: thus we get Agnostus agnostiformis (M'Coy),
Axinus axiniformis (Phillips), Dendwphyllia dendwphylloides
(Lonsdale), etc. Or again, a species has been named

THE RULES OF NOMENCLATURE 365

from its resemblance to some object, such as a sword or
an acorn, and the Latin name of that object has later
been applied to a new genus into which the species has
been transferred : thus we get Ensis ensiformis (S. V.
Wood) and Balanus balanoides (Ranz). Yet again, a
character unusual in its genus is taken for the trivial
name of a species which is afterwards transferred to
another genus in which that character is normal : thus a
spiny shell from the Chalk was taken by J. Sowerby to
belong to the genus Plagiostoma, and as the only spiny
species in that genus it was appropriately called Plagio-
stoma spinosum, but later it was transferred to the spiny
genus Spondyltts, but the name Spondylus spinosus, though
meaningless as a description, serves quite well as a
name.

The original Stricklandian code of 1841 proposed that
a name might be changed when it implies a false proposi-
tion which is likely to propagate important error. For
instance, Hyatt named an ammonite Mantelliceras
indianense, a name apparently implying that it was
found in the State of Indiana, whereas it actually came
from India. It might seem reasonable in this case to alter
the trivial name to indiense or indicttm, were it not that such
a change might lead to the idea that two distinct species
were in question. Really serious cases of this sort must
be so few that they can easily be kept in mind, and once
we depart from extreme cases we find individual judg-
ment introducing uncertainties ; consequently later zoolo-
gists have wisely dropped this proposal and agreed that
once a name has been published it may not be altered,
even by the author himself, except in correction of an
evident slip of the pen or misprint.

The Stricklandian code also proposed that where the
trivial name of a species was taken as the generic name
for a new genus in which that species was included, the

366 PALEONTOLOGY

trivial name must be altered, because they regarded the
exact duplication of generic and trivial names as bar-
barous. For example, Martin, in 1809, named a certain
brachiopod Anomites productus. J. Sowerby, in 1812, took
it as the type of a new genus, which he called Productus,
and changed Martin's species into Productus martini (quite
rightly according to the Stricklandian code). Later
zoologists have rejected this exception to the law of
priority, and the fossil which has long been known as
Productus martini must be called P. productus (Martin).

4. As knowledge increases and becomes more exact,
a later author may transfer a species from its original
genus to another. Thus Cheirurus insignis Beyrich, 1845,
is a synonym of Pavadoxides bimucronatus Murchison,
1839. The law of priority here settles that the earlier
trivial name must stand ; but as it was recognized that
Pavadoxides bimucronatus Murchison was not a true
Pavadoxides, it was transferred to Beyrich's new genus
Cheirurus, but the trivial name having priority over
Beyrich's name insignis, it stands as Chemirus bimucronatus
(Murchison), the author's name being placed in
parentheses as an indication that an alteration in the
generic name has been made. An alternative way of
expressing this is Cheirurus bimucronatus Murchison sp.

As an example of the importance of these distinctions,
which may at first appear an unnecessary refinement,
we may take the following case : James Sowerby, in
1818, gave to a lamellibranch the name Cardita deltoidca ;
in 1820 he gave to another the name Venericardia deltoidca.
In 1826 J. de C. Sowerby made a new genus, Pholadomya,
into which he moved the former of the above species.
In 1871, S. V. Wood transferred the latter of the
two from Venericardia to Cardita. Thus we "find that
Cardita deltoidea (J. Sowerby) is quite a different thing
from Cardita deltoidea J. Sowerby :

THE RULES OF NOMENCLATURE 367

Cardita deltoidea J. Sowerby = Pholadomya deltoidea
(J. Sowerby).

Venericardia deltoidea J . Sowerby == Cardita deltoidea
(J. Sowerby).

A wise precaution to avoid such possibilities of con-
fusion is never to use the same trivial name in genera at
all nearly related.

5. A later author may divide a genus into several
genera. The original name must then be retained for
one of these divisions, and, if possible, for that division
which was regarded by the author of the original genus
as most typical. With a view to this possibility the
founder of a genus should always name one species as
most typical : such a species is called the type-species of
the genus, or, more shortly, the genotype (see later). The
absence of a specified genotype from the original defini-
tion of a genus has often been a source of much con-
fusion when the genus has been subdivided later.

The subdivision of a genus often takes place in two
stages : one author divides it into sub-genera, and a later
author raises these to the rank of genera. To provide for
this possibility it is a rule that when a genus is divided
into sub-genera, one of these (the one including the
genotype) must bear the same name as the genus itself.
To distinguish the sub-generic name from the generic
name, the former is written with the addition of the
abbreviation s. sir. (senstt stricto, in the restricted sense).
Thus Hall and Clarke divided the old genus 0-rthis of
Dalman into a numbejr of sub-genera, to all of which
except one new names were given, such as Dalmanella,
Schizophoria, etc., but the one which included Dalman's
genotype was called Ovthis s. str. Under this scheme the
specific names of some well-known species became —

Orthis (Dalmanella) elegantida Dalman ;
Orthis (Schizophoria) resupinata (Martin) ;
Orthis (Orthis) callactis Dalman.

368 PAL/EONTOLOGY

Later authors have raised these sub genera to the rank
of genera, so that the same species now stand as —

Dalmanella elegantula (Dalman) ;
Schizophoria resupinata (Martin) ;
Ovthis callactis Dalman.

The printing of the author's name with or without
parentheses in the different cases should be noted
carefully.

6. A later author may unite several hitherto separate
genera into one. This is a much" rarer case than the last,
as the whole tendency of advancing classification is to
increase rather than to diminish the number of divisions
of all grades. In this case the application of the law of
priority leads to the rule that out of the names of the
several genera which are being united the earliest name
shall be retained as the name of the united genus.

The names of categories higher than genera cannot be
subject to the same rigid law of priority as generic and
specific names.

For families the rule is that the name of the family
is formed by combining the name of the type-genus with
the patronymic Latin termination idee. Thus the family
of which Olenus is the most typical genus is called
Olenidcc. If for reasons of priority it is found necessary
to change the name of the type-genus, the family name
must change with it. Thus if it is decided that the
generic name Goldius de Koninck has priority over the
usually accepted name Brontetis Goldfuss, then the family
name Bvonteidcz must also be altered to Goldiidce.

In the case of sub-families the name is similarly formed
by combining the root of the name of the type-genus
with the termination -ince. Thus the family Phacopidce,
typified by the genus Phacops, is divided into three sub-
families, named Dalmanitince, Phacopince, and PUrygo-
metopina, typified by the respective genera Dalmanites,

THE RULES OF NOMENCLATURE 369

Phacops, and Pterygometopus. It will be noted that the
genus Phacops is typical both of the whole family and of
one of the sub-families.

As there have been many cases where opinions have
differed as to the particular form which was denoted by a
certain name, and as neither printed descriptions nor
figures can always be interpreted with certainty, it is of
the utmost importance that the actual specimens which
an author had before him when founding a species should
be preserved for reference in case of doubt. These
specimens are called type-specimens or types, and they
are now always preserved with the greatest care by all
museums which possess any, being usually marked with
some special label — e.g., in the British Museum types are
indicated by a small circular green label.

Types are of several kinds :

If the author of a new species defines that species with
special reference to one particular individual, that in-
dividual is called the holotype of the species.

If, while specifying a holotype, he also refers to other
specimens in his original description, or figures them
along with the holotype, these additional specimens are
called paratypes.

If, however, he uses several specimens in his original
description without specifying one of them as the
holotype, these specimens are called syntypes (or
co-types).

If the syntypes of a species are subsequently dis-
covered to belong to two or more distinct species, the
original author, or (with his permission, or after his
death) any later author, may select one of them as the
type of the original species : this is called a lectotype.
The remaining syntypes, if not identical with other
species already named, may become holotypes of new
species at the same time. •

24

370 PALAEONTOLOGY

Types of a much lower degree of value are the
following :

Any specimen of a species coming from the same
locality and zone as its holotype or syntypes is a
topotype. Topotypes can be indefinitely increased in
number by collecting, whereas there can never be more
than one holotype or lectotype, and never more than a
few paratypes or syntypes.

A topotype recognized by the original author as be-
longing to this species acquires additional value thereby,
and is called a metatype.

A specimen of a species described or figured by the
original or any later author at some subsequent date to
the original establishment of the species is a plesiotype.

What a type-specimen is to a species a type-species or
genotype is to a genus: that is, it is a more certain
means of settling the application and extent of the generic
name than any verbal definition.

If the author of a genus defines one particular species
as typical of the genus, that is the geno-holotype.

If he only gives a list of the species which he considers
to belong to his genus, these are geno-syntypes.

Out of a series of geno-syntypes a geno-lectotype
may be chosen at any later date, either by the original
author of the genus, or (with his permission, or after his
death) by any subsequent author.

The collector of fossils who wishes to name his speci-
mens accurately will need to take much trouble in
hunting through works in which fossils are described and
figured. These works are of several categories.

First, we have monumental works, mostly of rather
early date, which set out to describe and figure all the
fossils (or all of some great group of fossils) found in
some one country or State.

Secondly, there are numerous smaller works, some-

THE RULES OF NOMENCLATURE 371

times separately published, sometimes forming parts of
scientific journals, in which only some small groups of
fossils, or the fossils from some one locality or district,
are described.

Thirdly, there are various monographs, such as those
published by the Palaeontographical Society of London,
which aim at a complete description (with revision of all
previous publications on the subject), either of some
great group as the Brachiopoda, for all geological
systems, or for some one system (as the Cretaceous
Lamellibranchs or the Carboniferous Trilobites), or,
more rarely, the whole fauna of some one system or
formation. Such monographs, if properly prepared, are
of enormous value to students, as they sum up all that is
known up to the date of their publication, and only very
special investigators need go behind them to earlier
works. These three categories are fairly distinct, but
it is not always possible sharply to separate them.

Fourthly, there are catalogues and indexes which,
without themselves describing any new genera or species,
enable students to find with the least labour those works
where they are described.

Some of these monographs, etc., have already been
noted in the short bibliographies appended to each
chapter. It remains to list a few of the works of more
general kind. Fuller bibliographies of particular groups
of fossils will be found in some of these.

General Bibliography.

I. — INDEXES.

BRONN, H. G.— Index Palaeontologicus (1848-49).
Records every species of fossil plant and animal named
up to date of publication, with references to author,
synonymy, geological age,^etc.

372 PAL/EONTOLOGY

MORRIS, J. — British Fossils, 2nd edn. (1854). ^e~
cords all British species up to date of publication, with
references.

SHERBORN, C. D. — Index Animalium, Vol. I. (1902).
Complete index to all animal genera and species, recent
and fossil. The published volume covers publications
from 1758 to 1800. Vol. II. will extend to 1850.

ZOOLOGICAL RECORD. — Annual publication by the
Zoological Society of London, including lists of all new
genera and species of animals published during the year.

PAL^ONTOLOGIA UNIVERSALIS. — A series of photo-
graphic reproductions of type-specimens and their original
figures and descriptions (chiefly of species published
before 1865).

FOSSILIUM CATALOGUS, edited by Professor Freeh, of
Breslau (Berlin, 1913 — ). — A series of lists of species
compiled by specialists, with geological and geographical
range, synonymy, and references, thus replacing Bronn's
Index, piece by piece. Lists so far published include
Cretaceous Corals (by J. Felix), Palaeozoic Starfish (C.
Schuchert), Tertiary Anisomyarian Lamellibranchs (W.
Teppner), Devonian Ammonoidea (F. Freeh), Triassic
Cephalopods (C. Diener), Triassic Dinosaurs (F. de
Huene), Stegosauria (E. Hennig), Lycopodiales and
Equisetales (W. Jongmans), and Juglandaceae (K. Nagel).

II. — LARGE GENERAL TEXTBOOKS.

BERNARD, F. — Elements de Paleontologie, pt. i.
(Paris, 1893).

FISCHER, P. — Manuel de Conchyliologie (1880-87).
Deals with recent and fossil Mollusca and Brachiopoda,
and is still the best work on Lamellibranchs and
Gastropods.

NEUMAYR, M. — Die Stamme des Thier-reiches (1889).
A treatise on phylogeny, left very incomplete by the

THE RULES OF NOMENCLATURE 373

author's early death. Deals only with Protozoa, Coelen-
terata, Echinoderms, and Brachiopods.

NICHOLSON, H. A., AND LYDEKKER, R. — Manual of
Palaeontology (1889), Vol. I.: Invertebrata, Vol. II.: Verte-
brata and Plants. Still useful, though out of date in
certain sections.

ZITTEL, K. A.— (i) Handbuch der Palaontologie,
4 Vols. (1876-93) ; also French translation by Ch.
Barrois (1883-94). The most complete textbook, but
suffering from the lapse of time since its publication.
(2) Grundzuge der Palaontologie (1895, new edition,
1910-11), also English translation by C. Eastman, Vol. I. :
Invertebrata (1900, 2nd edn. 1913), Vol. II.: Vertebrata
(1902). (3) History of Geology and Palaeontology,
translated by M. M. Ogilvie-Gordon (Contemporary Science
Series, 1901). The chapters on Palaeontology and Strati-
graphical Geology are indispensable to any student wish-
ing to understand the history of research and the work of
the great pioneers.

III. — MONOGRAPHS OF EXTENSIVE NATURE.

BARRANDE, J. — Systeme Silurien du Centre de la
Boheme (1852-99). A series of very detailed and mag-
nificently illustrated monographs on the fossils of the
Lower Palaeozoic rocks of Bohemia. Barrande's own
work covers the Trilobites, Mollusca, and Brachiopods;
and after his death other authors have added volumes on
Graptolites, Corals, Bryozoa, and Cystids.

DESHAYES, P. G. — Description des Coquilles Fossiles
des Environs de Paris, 2 vols. and Atlas (1824-37) :
Cainozoic Mollusca.

D'ORBIGNY, A. D. — Paleontologie Fran9aise (1840-55,
with later volumes by other authors). An attempt, never
completed, to figure and describe all the fossils of France.

374 PALAEONTOLOGY

It covers Jurassic and Cretaceous Corals, Echinoderms,
Bryozoa, Brachiopoda, Mollusca and Plants, also Eocene
Echinoids.

GOLDFUSS, G. A. AND MONSTER, G. — Petrefacta Ger-
manise, 2 vols. (1826-40). An equally ambitious design
for Germany, limited in execution to Sponges, Corals,
Echinoderms, and Mollusca.

HALL, J.— Palaeontology of New York (1848-98). A
series of elaborate monographs on the Palaeozoic Fossils
of New York State.

LAMARCK, J. B. DE M. — Histoire Naturelle des Ani-
maux sans Vertebres (1815-22).

SOWERBY, J., CONTINUED BY SOWERBY, J. DE C.

Mineral Conchology of Great Britain (1812-46). Pub-
lished in monthly parts, with interruptions, its object was to
figure British fossil shells, i.e., Mollusca as then understood
(including Foraminifera, Cirripedia, Annelida, Brachio-
poda, and, by mistake, one Coral). In all about 1,250
species were described, and figured on 648 plates, en-
graved on metal and coloured by hand with great faith-
fulness to the natural appearance. There is no system
in the sequence of species, and as British stratigraphy
underwent its development during the publication the
geological age is much better treated in the later than in
the earlier parts. Many of the types described were
unfortunately derived fossils from the Glacial Drift.

The above lists might be extended indefinitely, but
their object is to give the student a first general idea of
the sort of material available for the purpose of identify-
ing the fossils in his collections, and for extending his
knowledge in any department that may specially interest
him.

\

APPENDIX I
DIVISIONS OF GEOLOGICAL TIME

DIVISIONS OF GEOLOGICAL TIME 377

DIVISIONS OF GEOLOGICAL TIME. •

The student of Palaeontology constantly requires to
refer to the names which have been given to the divisions,
great and small, into which the fossiliferous rocks and
the time of their deposition have alike been divided.
These are here given in tabular form for convenience of
reference. For the grounds on which their limits are
fixed the reader is referred to textbooks of Stratigraphy,
or he will find a reasoned statement in a paper by
MM. Munier-Chalmas and de Lapparent : "Note sur la
nomenclature des terrains sedimentaires," Bull. Soc. Geol.
France, 3e ser., t. xxi. (1893), P- 43$-

Some possible sources of confusion must be pointed
out. First, the necessity for a double series of terms —
more or less local stratigraphicaRerms (such as Wenlock
Limestone, Carboniferous Limestone, Gault) and terms
of world-wide, or at least international, application (such
as Cambrian, Visean, Cenornanian). Both have their
own usefulness, but the attempt to use terms of the first
category in place of those of the second has led to much
confusion in correlation ; while forgetfulness of the fact
that the second category should apply, as far as possible,
over the whole world has led geologists to coin some
unnecessary synonyms. The limits of the first set of terms
are determined by marked changes in the character of the
sediments laid down in a limited area; those of the
second should be determined primarily by widespread
changes of fauna or flora, and secondly by great changes
in the distribution of land and sea (the biological and
geographical changes being intimately related).

The tables on pp. 378-380 must not be taken as fixed
and settled. Geologists are not in complete agreement
as to the relative importance of many divisions : what is
a series to one is only a stage to another, and nomencla-
ture varies accordingly. A few synonyms, etc., are given
here, in descending order :

Montian is included in Danian by some.
Aturian = Maestrichtian + Campanian .
Senonian = Aturian + Emscherian.

378

PALEONTOLOGY

Group
Era.

System
Period.

Series
Epoch.

Stage
Age.

Local Stratigraphical
Terms (chiefly
British) and Synonyms.

I

Holocene

—

—

Pleistocene

—

Glacial Drift, River
gravels, etc.

Pliocene -j

Sicilian
Astian
Plaisancian

Red Crag
Coralline Crag

CAINOZOIC

Miocene

Pontian
Sarmatian

Tortonian
Helvetian
Burdigalian
Aquitanian

fCongeria-beds,
J etc. , of Caspian
j facies in S.E.
I Europe

(Swiss Molasse,
Faluns of Tour
aine

Oligocene

Stampian
Sannoisian

—

PAL^OGENE ..."
(Nummulitic)

Eocene

Bartonian
Lutetian

Londinian
(Ypresian)
Thanetian
Montian

Calcaire Grossier
(Paris)
London Clay

Upper (
Cretaceous «
(Senonian) |

Danian
Maestrichtian
Campanian
Emscherian

\Upper White
/ Chalk

MESOZOIC

CRETACEOUS ....

Middle /
Cretaceous I
(often
included 1
mainly in
Upper Cret.) v

Turonian
Cenomanian

Albian

Middle Chalk
Lower Chalk, Chalk
Marl, Upper
Greensand
Gault

Lower
Cretaceous "

Aptian
Barremian
Hauterivian
Valanginian

Lower Greensand
/Freshwater equiv-
\ alent = Wealden
Upper Volgian of
Russia

DIVISIONS OF GEOLOGICAL TIME

379

1

Group
Era.

System
Period.

Series
Epoch.

Stage
Age.

Local Stratigraphical
Terms (chiefly
British) and Synonyms.

,

Aquilonian

Freshwater equiva-
lent = Purbeck

beds

i

Portlandian

—

Bononian

—

Upper
Jurassic •
(Oolitic)

Kimmeridgian
Sequanian
Argovian

j-Corallian

Divesian
Callovian

| Oxford Clay

Bathonian

Great Oolite

O

JURASSIC

Vesulian
Bajocian

1 Inferior Oolite

3

Aalenian

.

0

I

Yeovilian

\Toarcian; Upper

C/)

Whitbian

/ Lias

W

Lower

Domerian

Middle Lias (Marl-

S

Jurassic -

stone)

(Liassic)

Charmouthian

I

(Carixian)
Sinemurian

h Lower Lias

Hettangian

)

Rhastian

Norian

\ Non-marine equiv-

Carnian

/ alent = Keuper

TRIASSIC

Ladinian

Muschelkalk of

Germany

Anisian

~\ Non-marine equiv-

Scythian / alent = Bunter

(

Thuringian

Magnesian Lime-

stone

u

KH

PERMIAN

Penjabian

Saxonian *

ttIO
UN

Artinskian

Autunian *

—

^g<

(

Uralian

Stephanian *

; W ^

Moscovian

Westphalian*

British Coal Meas-

<*{

CARBONIFEROUS J

•

ures

PH

1

Dinantian \

:Visean
Tournaisian

\Carboniferous
j Limestone

* These are not names of stages, but equivalent names for the non-marine equivalent series, th<
Coal Measures and similar formations, which are of greater importance in these two systems thar
in any others.

PALEONTOLOGY

Group
Era.

System
Period.

Series
Epoch.

Stage
Age.

Local Stratigraphical
Terms (chiefly
British) and Synonyms.

JJ

Upper

Fammenian

Upper Old Red
Sandstone

PH O

I

Frasnian

£g<

DEVONIAN

Mid'dle |

Givetian
Eifelian

~

§5

f

Coblentzian

Lower

Gedinnian

Lower Old Red

PH

I

Sandstone

[

Downtonian

—

Upper Ludlow beds,

etc.

SILURIAN ... -f

Salopian

—

Wenlock and Lower

u

0
N

(Gothlandian)

Valentian

—

Ludlow beds
Llandovery beds

1

f

Ashgillian
Caradocian

—

Bala beds

ORDOVICIAN ...-,'

Llandeilian

—

d.

|

Llanvirnian

tf

I

Skiddavian

—

Arenig beds

Q

Tremadocian

_

_

Potsdamian

—

Lingula Flags

CAMBRIAN

Menevian

(Acadian)

Georgian

—

—

Neocomian=Hauterivian + Valanginian, and is sometimes used to include
also Barremian and even Aptian.

Urgonian is the name of a special facies, partly Barremian, partly Aptian.

Tithonian = Portlandian + Aquilonian approximately, as developed in Alpine-
Mediterranean region.

Volgian = from Portlandian to Barremian, as developed in Russia.

Bononian is in England generally divided between Portlandian and Kim-
meridgian.

Corallian is a facies name of variable age, roughly Argovian and Sequanian.

Oxfordian is used in several senses ; it covers Divesian with either higher
stages up to Sequanian or part of Callovian below.

Bathonian as originally denned included Vesulian, and Bajocian included at
least a part of Aalenian, and these terms are still used in the wider sense.

Toarcian = Yeovilian + Whitbian .

Pliensbachian = Domerian + Charmouthian ; and Charmouthian was origin-
ally synonymous with Pliensbachian.

Charmouthian has recently been divided by Mr. Buckman into Hwiccian,
Wessexian, and Raasayan, and Sinemurian into Deiran, Mercian, and
Lymian (see pp. 388-9).

DIVISIONS OF GEOLOGICAL TIME 381

Werfenian = Scythian.

Anthracolithic = Permian 4- Carboniferous.

Avonian-Dinantian practically.

Silurian is used in three senses: (i) as here = Gothlandian, the
usual English use; (2) as=Gothlandian + Ordovician, the
usual French and German and former ^English use : this use
may always be assumed when ' ' Lower " and " Upper Silurian ' '
are spoken of ; (3) as = Older Palaeozoic, not a frequent use.

The division of stratified rocks into zones — i.e., divisions
less than stages based on palaeontological characters alone,
and extending, if not over the whole world, far beyond
the limits within which lithological distinctions remain
constant — was initiated by Oppel for the Jurassic system
about sixty years ago. It has since been applied to
most other systems, but no other system has proved
susceptible to such detailed application of the method.
As Oppel's original zones proved capable of division to
a greater or less degree, Mr. S. S. Buckman proposed
the term hemera as the name of the smallest possible
time-division which palaeontological facts make possible —
it is the period of acme of a particular species (or
occasionally of a genus). The term zone has since been
regarded by many as the rock-division corresponding
to a hemera ; but Dr. Lang and others consider that the
zones defined by Oppel must continue to stand, and
subdivisions made in them corresponding to hemerae
must be called sub-zones. Accordingly, in the table of
Jurassic zones, both Oppel's original zones and the
latest recognized hemerae are given. In quoting the
hemerae (or zones) it is usual to give only the trivial
name of the index fossil, thus : Uarmatum zone, obtusum
hemera (or hemera obtusi, using the genitive case).

The fossils that have proved of the greatest value as
zone-fossils are Ammonoids and Graptolites ; but
Brachiopods, Trilobites, Foraminifera, and Lamelli-
branchs (in order of decreasing value) also serve ; and
isolated species of other groups, e.g., Crinoids, may often
be of great use. In freshwater Cainozoic deposits
Mammals are of the highest value.

It should be unnecessary to add that these tables of
zones are given for reference only, and by no means as an
exercise in memorizing !

382

PALEONTOLOGY

1

B

;

ll

1

1

'•S "ft.

§ -~ to

f

JH

-2

'S

•S

\

\

\

II

\

1 i-ll

be

K

^ ^

to ^2

to ^ ^ £*

b

«

?i -^

^* 5fi

-£i a

K

M

6

It

11

1 §11

0

2.5

00

$5^3

rt

•

\ ^

\

j>

<

*c

\

1

\

-Ss

\

"s

\

^

z

j^

53

•^

a

J

**

13 <3

&

*

a

LEPIDO

|

-a .
O

\

1

marginata

1

|
^

2

I'i

\

0

ffi
O

1

p

ll

^

coo

Jl

h

2

<o

"3

S

s

•3

•£|

D

\

v>

|

<o

vf-Si

vT^

Co

V)

js

~ ??

"^5

<o '^

D
fe

1

%

•1

1 2

1 1

s^i

"BI

It

^

^ ***-

Q 2

•g s

1

•§

|

5 §o

^^

ll

?:

8

•5

s

2

1

;

I

•£

z

!

£

:

H

H

H

:

H

W

\

<
n

p

O?

C*
<

P

z

z

<

z

<

H

BURDIGA

5

MIDDLE

M
H

STAMPIA

SANNOIS

BARTONI

%

OH

|

1

3N3DOIIM

3N3O

•oono

3N30O3

DIVISIONS OF GEOLOGICAL TIME

H

\

\

3

1

s

M

M

as

'S

O

,g

0

5

13

K

1

\

\

0
M

\

1

\

1

\

\

\

\

\

s

'i

'"§

1

g"

2

§

Ji

•2

*

;S

a

1

$

•2

1

s

1

i

I

\

\

\

ITOIDES.

\

1

1

\

\

\

\

CS

^>

^

0

•S

i

^

•§

•S

K

.8

^

"8

§0

§

§

i

.g

•§

\

\

z

§

1

s

\

\

\

1

\

\

\

o

H

o
8

o

.^>

^

03

^

,°*

v>

|

to

»

0

g

1

1

1

1

1)

E

1

|

••§

1

« "3

i

5

^

§

5

^

"K

•

;

z

z

:

;

^

<

•aj

w

\ LOWER LUTETIAI

YPRESIAN

THANETIAN ...

MONTIAN

NVINVQ

MAESTRICHTIAN

(DORDONl

CAMPANIAN ...

EMSCHERIAN

\ TURONIAN ...

UPPER CENOMANI

LOWER CENOMA>

ALBIAN

APTIAN

BARREMIAN ...

3N3003

PALEONTOLOGY
2. CRETACEOUS ZONES.

Stages.

Ammonoid Zones of the South of
France, after Kilian, Grossouvre,
and Jacob.

Zones of England,
North France, etc., after
Barrois, Jukes-Browne, etc.

DANIAN

Hercoglossa [Nautilus] danica

—

MAESTRICHTIAN J

Parapachydiscus neubergicus
Bostrychoceras polyplocum

—

CAMPANIAN ... J

Hoplites vari
Mortoniceras delawarensc
Placenticeras bidorsatum

Ostrea Innata
Belemnitclla mucronata
Actinocamax quadratus

EMSCHERIAN ... -!

Placenticeras syrtale
Mortoniceras texanuin
Mortoniceras emscheris
Barroisiceras h aberfellneri

Marsupites testudinarius
Micras ter cor-anguin um

Micraster cor-testudinarium

TURONIAN ...-j

A canthoceras deverianum
A canthoceras ornatissimum
A canthoceras bizett
Mammites nodosoides

Heteroceras reussiamim
Holaster planus
Terebratulina lata
Rhynchonella cuvieri

CENOMANIAN... -I

A canthoceras rotomagense
Mantelliceras mantelli
Mortoniceras inflatwn

Holaster subglobosus
Schloenbachia varians
Mortoniceras inflatum

ALBIAN ... -j

Mortoniceras hugardianum
Hoplites dentatus
Hoplites tardefurcatus

Hoplites lautus
Hoplites interruptus
Douvilleiceras mammillatu'ni

APTIAN ...j

Dovvilleiceras nodosocostalum
Douvilleiceras subnodosocosta-
tum
Oppelia nisus
Parahoplites deshayesi

Parahoplites deshayesi

BARREMIAN ... -j

Heteroceras astierlanum
Pulchellia pulchella
Parahoplites angulicostatus
Desmoceras sayni

Russian Zones, after
Pavlov.

Simbirskites speetonensis
Simbirskites subinversus

HAUTERIVIAN

Crioceras duvali
Acanthodiscus radiatus

Polyptychites polyptychus
Polyptychites keyserlingi

VALANGINIAN

Saynoceras verrucosum
Kilianella roubaudiana
Thurmannia boissieii

Craspedites stenomphalus
Craspedites spasskensis
Polyptychites gravesiformis

DIVISIONS OF GEOLOGICAL TIME 385

3. ZONES OF THE UPPER JURASSIC.

Oppel's Zones
(1856-58).
Approximate
Equivalence.

ij£
%£

<a "
p

<o>

Stages.

Heinerae (Burkman, Salfeld,
etc., 1898-1915).

AQUILONIAN

Craspedites fragilis

Trigonia
gibbosa

PORTLANDIAN

Perisphinctes gigantens
Perisphinctes pseudogigas
Perisphinctes gorei
Perisphinctes eastlccottensis

Pterocera
oceani

BONONIAN

Perisphinctes pectinatns
Virgatites pallasianus
Virgatites miatschkovensis
Gravesia irius
Gravesia gravesiana

Astarte
supracorallina

White
Jura

KlMMERIDGIAN

Anlacostephanuspseudo-mutabilis
Aulacostephanus yo
Rasenia mutabilis
Rasenia cymodoce
Pictonia baylei

Diceras

anetiniim

e
5
7
P

SEQUANIAN
(Age of

Amceboccras)

Ringsteadia pseudocordatus
Perisphinctes decipiens (and
Amceboceras serratwn]
Perisphinctes wartce

Cidarisflori-
gemma

Ammonites
biarmatus

a

ARGOVIAN
(Age of
Cardioceras)

Perisphinctes martelli
A sp idoceras biarma turn
Cardioceras vertebrah (and
Cardioceras co rda turn )
Cardioceras scarburgense

DlVESIAN

(Age of
Quenstedtoceras)

Quenstedtoceras gregarium
Quenstedtoceras vertumnus
Creniceras renggeri
Quenstedtoceras lamberti

* Quenstedt divided the whole Jurassic System into three— the Black Jura, Brown Jura,
and White Jura, for which Oppel's equivalent terms were Lias, Dogger, and Malm. Each
of these three w,\s divided into six divisions, marked by the Greek letters a, £, y, 8, e, and £.
The whole of the White Jura and most of the Brown Jura come in the Upper Jurassic ; the
rest of the Browa and all the Black Jura in the Lower Jurassic.

385 PALAEONTOLOGY

ZONES OF THE UPPER JURASSIC— Continued.

Oppel's Zones "S^< '

(1856-58). ;, |?o

Approximate 5 „

c. Hemerze (Buckman, Mascke,
StaSes- etc., 1898-1915).

Equivalence.

5

A mmonites

Brown

Peltoceras athleta

athleta Jura

Cosmoceras duncani

A mmonites

>

Cosmoceras castor

anceps

^!rLOV ^N \ Erymnoceras coronatum

Cosmoceratids) s2K2SiS2S2j

A mmonites

— ^ — . — .

Proplanulites koenigi

macrocephalus

Macroccplialitcs macrocephalus

Terebratula

Clydoniceras discus

lagenalis

Epitliyris marmorea *

Terebratula

BATHIAN

Omiihclla digona *

digona

e

(Age in which

Ornithella digonoidcs *

Ammonites

Epithyris bathonica *

are rare)

Oppelia waterhousci
Macrocephalites morrisi

Perisphinctes gracilis

e

Ostrea acuminata |

Ammonites
parkinsoi.i

VESULIAN
(Age of
Parkinsonia)

Oppelia fusca
Zigzagiceras zigzag
Parkinsonia schloenbachi
Strigoceras truellii

5

Garantiana garantiana

5

Strcnoceras niortcnse

Ammonites
humphriesianus

7

BAJOCIAN

Teloceras blagdcni
Stepheoceras
Stephanoceras
Stemmatoccras

Ammonites

(Age of
Sonninians)

Otoites sauzei
Witchellia
Shirbuirnia

sauzci

post-discitcs

;

Hypcrlioceras discites

* Brachiopods

f Lamellibranch.

DIVISIONS OF GEOLOGICAL TIME

387

4. ZONES OF THE LOWER JURASSIC.

Zones (Oppel,
1856-58).

Quenstedt
(1858).

Stages.

Henierce (Buckman, etc., 1913-19).

Ammonites
murchisoni

Brown
Jura

0

AALENIAN
(Age of oxynote

Ludwigella concava
Brasilia bra df or den sis
Ludivigia murchisonce
Ancolioceras

Trigonia
navis

a

Hildoceratids)

Tmetoceras scissum
Cypholioceras opaliniforme
Pleydeliia aalensis

Ammonites
torulosus

Black
Jura

YEOVILIAN
(Age of

Dumorticria moorei
Catulloceras
Dumortieria sp.
Hammatoceras

Ammonites
jurensis

r

Dumortierians
and
Grammoceratids)

Phlyseogrammoceras dispansum
Pseudogrammocems struckmanni
Pseudogrammoceras pedicum
Haugia eseri
Grammoceras striatulum

Posidononyma
bronni

e

WHITBIAN
(Age of
Dactyloids
and
Harpoceratids)

Haugia variabilis
Lillia lilli
Collina brauniana
Peronoccras fibulatum
Frechiella subcarinata
Ovaticeras pseudovatum
Harpoceras falciferum
Harpoceras cxaratum
Dactylioceras tenuicostatum
Tiltoniceras acutum
Harpoceratoid

Ammonites
spinatus

Paltopleuroceras spinatum

A mmonites
margavitatus

d

DOMERIAN

(Age of
Amaltheids)

Amaltheus la vis
A maltheus gibbosa
Segucnziceras algovianum
Caloceras acanthoides
Hildoceras boscense
Grammoceras fieldingi

388

PALEONTOLOGY

ZONES OF THE LOWER JURASSIC— Continued.

Zones (Oppel,

1856-58).

v^z.

I!

Stages.

Hemerae (Buckman, Trueman,
Thompson, etc., 1910-19).

O1

Black

Oistoceras

A mmonites
davoci

Jura

HWICCIAN

JEgoceras desdalocosta
Deroceras davoei
Amblycoceras byevilobatum

7

(Age of
Liparoceratidse)

JFgoceras lattscosta
Beaniceras

Acanthoceras carinatum

Ammonites

Liparoceras cheltiense

ihnv

loex

Acanthopleuroceras valdani

Tragophylloceras ibex

A canthopleuroceras ellipticum

Ammonites
jamesoni

7

WESSEXIAN
(Age of
Polymorphic! ae)

Uptonia bronni
Pla typleuroceras
Polymorphites trivialis .
Uptonia jamesoni

Codoceras pettos

Polymorphites peregrinus

Phricodoceras

A mmonites

Deroceras leckenbyi

armatus

RAASAYAN

7

(Age of

Echioceras applanatitm

Deroceratidse Echioceras macdonnelli

A mmonites

and

Echioceras raricostatoides

raricostatus

P

Echioceratidae)

Deroceras bispinigerum
Bifericeras subplanicosta

ist Echioceras '

Ammonites
oxynotus

P

DEIRAN *
(Age of Oxycone
Arietids)

Oxynoticeras oxynotum
JEtomocevas simpsoni
Gagaticcras

* The existence of several other hemerae in the Deiran and Mercian is indicated by Mr.
Buckman, but they are not definitely established, and are therefore omitted. See Q uart.
Journ. Geol. Soc., vol. Ixxiii., pp. 257-327(1918).

DIVISIONS OF GEOLOGICAL TIME

389

ZONES OF THE LOWER JURASSIC— Continued.

Zones (Oppel,
1856-58).

Quenstedt
(1858).

Stages.

Hemerae (Buckman, Tutcher,
etc., 1913-19).

Ammonites
obtusus

Black
Jura
(9

MERCIAN *
(Age of
Catagenetic
Arietids) •

A steroceras stellare
Xipheroceras planicosta
A steroceras obtusum
A steroceras brooki
A rietites turneri f

Pentacrinus
tube} culatus

a

LYMIAN

(Age of
Anagenetic
Arietids)

Microderoceras bircJii
Arnioceras semicostatum

Agassiceras sauzeanum
JEtomoceras scipionianum

Coroniceras gmuendense
Coroniceras UiMandi
Coroniceras rotiforme
Vermiceras -conybeari

Ammonites
geometricus

Ammonites
bucklandi

Ammonites
angulatus

a

HETTANGIAN
(Age of
Caloceratids)

Schlotheimia angulata
Alsatites liasicus
Waehneroceras megastoma

C aloe eras johnstoni
Psiloceras planorbis

Ostrea lias si ca J
Pleuromya tatei J
Volsella [Modiola] langportensis I

Ammonites
plan or bis

" Bone-bed "
(with Rhaetic)

* The existence of several other hemerae in the Deiran and Mercian is indicated by Mr.
Buckman, but they are not definitely established, and are therefore omitted. See Quart.
Journ. Geol, Soc., vol. Ixxiii., pp. 257-327 (1918).

t There is some uncertainty as to the position of this hemera ; it may be below

I Lamellibranchs: no Ammonites are known outside the Alpine-Mediterranean province
before the piano? bis hemera.

390

•PALAEONTOLOGY

5. TRIASSIC ZONES.

This scheme of zones was drawn up by Suess,
Mojsisovics, Waagen, and Diener in 1895. The names
of the Stages have been somewhat modified since, and
as given here represent Dr. Diener's latest usage.

Stages .

Zones.

Famous Localities.

RHJETIC

Pteria [Avic^lla] contorta

Aust Cliff (Gloucester-
shire)

NORIC

Sirenites argonaut a
Pinacoceras metternichi
Cyrtopleurites bicrenatiis
Cladiscites ruber
Sagenites giebeli

VHallstadt (Austria)

CARNIC ... -|

Tropites sub-bullatus
Trachyceras aonoides
Trachyceras aon

St. Cassian (Tyrol)

LADINIC ... J

Protrachyceras archelaus
Dinarites avisianns
Protrachyceras curionii

| Muschelkalk of Thur-
f ingia, Lorraine, etc.

ANISIC T
(Dinarian, -I
Virglorian) (

Ceratites trinodosus
Ceratites binodosus
Stephanites superbus

—

SCYTHIC

Flemingites flemingianus
Flemingites radiatus
Ceratites normalis
Proptychites trilobatus
Proptychites lawrencianus
Gyronites frequens
Octoceras woodwardi

VWerfen (Austria)

\ Ceratite Marls and
V Salt Range (Pun-
J Jab)
Julfa (Armenia)

DIVISIONS OF GEOLOGICAL TIME
6. UPPER PALAEOZOIC ZONES.

39'

The zoning of the Upper Palaeozoic rocks is in a much less
satisfactory state than that of either the Lower Palaeozoic or the
Mesozoic. Professor Freeh, of Breslau, in 1898, proposed a number
of broad gomatite-zones (rather stages than zones, really) for the
Devonian system, and Professor Haug, of Paris, did the same for
the Carboniferous and Permian. These zones, as modified by
Professor Haug in his Trait e de Geologic (1911), are here given :

Systems.

Series.

Stages.

Ammonoid Zones.

Famous Localities.

2

THURINGIAN

-

20. Cvclolobus oldhttini

Chideru (Salt Range,
Punjab)

'S, *

SAXONIAN ...

-

ig. Xenaspis ca.rbona.rius

Virgal (Salt Range)

*

"• I

ARTINSKIAN... !

|

1 8. Waagenoceras tuoj-
sisovicsi
17. Medlicottia artiensis

Amb (Salt Range)
Ural Mountains (Russia)

(fi

URALIAN

—

16. Gastrioceras tnaria-
num

Ural Mountains (Russia)

)NIFEROtJ

MOSCOVIAN ... <

- (

15. Gastriocetas nsteri
14. Glyphioceras striola-
tum -^

{Myachknvo, ne ir Mos
cow (for Crinoid>)
The British coal-
measures are the non-
marine equivalent

»

<£
<
u

DlNANTIAN ...-!

Visean
Tournaisian

13. Goniatites striat-us

12. Pericyclus princess
and Aganides rota-
te* JUS

Bristol, Clitheroe (Lanca-
shire), Tournai (Belgium)

Famennian...-!

ii. Gonioclymenia pla,>ia
10. Clvmenia annulata
9. Chiloceras curvispina

Sauerland (Westphalia)
Kellerwald

£

Ul'PfR ...-I

Frasnian .. -\

8. Gephyroceras mtu-
mescens
7. Gephyroceras hoen-
ing/iausi

Btidesheim(Kifel), Naples
(U.S.A.)

<j
Z

o
>

Givetian ...<

6. Menecerat decheni
5. Anarcestes denck-
inanni

Givet (Ardennes)
Paffrath (near Cologne)

w

Q

Eifelian ...-[

4. Agoi/iatites occultus
3. /•/ narces*es subnau-
tilimis

Eifel (Rhine)
Couvin (Ardennes)

\

Coblentxian

2 . /?£ 0« /« tiles fide Us

Hunsriick slates with py-
ritic crinoids, etc.

LOWER ... -j

Gedinnian ...

i. 7 ornoceras itie.rpec-

t at 11 HI

—

392 PALEONTOLOGY

7. LOWER CARBONIFEROUS (DINANTIAN
OR AVONIAN) ZONES.

A scheme of zoning for the Carboniferous Limestone
of the Bristol district was developed by the late
Dr. Vaughan, and published in 1905. It proved
applicable over a wide area in South West England and
Wales and, with modifications, to the North of England
and Belgium. The zone-fossils being shallow water
Corals and Brachiopods, it is doubtful if the zones can
hold good over a much wider area. The table on p. 393
gives Vaughan's zones, and those established by
Professor Garwood in the North of England.

8. FORAMINIFER ZONES OF CARBONI-
FEROUS AND PERMIAN.

Fusulina kattaensis = base of Saxonian.
Fusulina verneuili= Artinskian.
Schwagenna princeps = Upper Uralian.
F'usulina longissima = Lower Uralian.
Fusulina cylindnca = Moscovian.
Fusulinella struvii = Visean.

These zones appear to cover the complete range of the
sub-family Fiisulinina.

DIVISIONS OF GEOLOGICAL TIME

393

T3 </, '

all

||

a

l

6

g

Sub-Zones.

Productus latissimus
Lonsdaleia florifo
Cyathophyllum mu
soni

N ematophyllum n
and Cyrtina carbo

Gastropod beds

Daviesella carinata
Camarophoria isorfy

Seminula gregaria
Solenopora

v , •

* Y '

^^-^

^-"

O

14

.

O

|

§0^ 5

1

•|

t/>

a G o

«

•^^

c

1 5

«P

£

N

^

S o, S

S

-^

2 3

.3

s>*>

s

^ s .*

5

•a

}

'o "^^

i

f

Sub-Zones.

Cyathaxonia
Lonsdaleia flori-
formis
Dibunophyllum 6

Productus cor-
rugato-hemi-
sphericus
Caninia bristo-
lensis

'•e-

§

3

1
|

Caninia cornuco-
pia

4 1
••8 ^

N^S1-2

"C ;§ K "^ ~y.

CO N° H

N M

N

M

N M

N M fl

QQ Q

CO C/)

O

O

N N

M W «

v _t

^ J

^ j

^ _T j

^

a

T

*

r

FOY

1

%

\

t^i

:

1 .3

11

1

i , a

2

II

1 1

<x>

N ! J

1

•2 'Sk

;§

O

-s:

1

>B

5* -S

^o $t

J|

2

o

I

§

CO

ft

N

0

KVHS.A

KV.SW™

394

PALAEONTOLOGY

9. UPPER DEVONIAN AMMONOID ZONES.

A much more detailed scheme of zones, comparable
with the zones of the older Palaeozoic or Mesozoic, has
lately been established by Dr. Wedekind of Gottingen
for the Upper Devonian of Germany.

Stages.

Zones.

FRASNIAN

/ VII.

FAMENNIAN ..

VI. /

Gonioclymcnia

V.'-! Clymenia cf. undttlata

\Clymenia lavigata

IV/3. Postprolobites yakelowi and Clymenia ttnui ,

cost at a
IVa. Clymenia crassicosta and Prolobites delphinns

HI/3. Clymenia invohtta
Ilia. Tornoceras sandbergeri

11)3. Cheiloceras and Dimeroceras
Ha. Cheiloceras

Id. Crickites holzapfeli and Manticoceras crassiim
I-y. Manttcoceras carinatum and Manticoceras cot'

datum

Ij8. Manticoceras nodulosum
la. Prolecanites lunulicosta

Zones la— 6 correspond to 7 and 8 of the preceding list ; I la and /3to 9 ; Ilia tc
V. to 10 ; VI. and VII. ton.

DIVISIONS OF GEOLOGICAL TIME

395

10. SILURIAN GRAPTOLITE ZONES.

After Lapworth, Elles and Wood.

Series.

Stages.

Graptolite Zones and Sub-Zones. Famous Localities.

DOWNTONIAN

No Graptolites known Ludlow (Shropshire),

bone-beds, etc.

/

£

36. Monograptus leintwar-

Aymestry (Shropshire),

o

dinensis non-graptoiitic Con-

T3
3

cliidium limestone

h-1 ,
I*

0)

J

35. Monogmptus tumescens \ Ludlowdistrict(GraP-
34. Monograptus scamcus 1 tolifcs and Nautili

33. Monograptus niissom •. . \
32. Monograptus vttlgans ) ea> tc''1

SALOPIAN ...

J_i

O

r^

31. Cyrtograptits Inndgreni
30. Cyrtograptus rigidus
29. Cyrtograptus iinnars-

rThe uppermost zone
is the approximate
horizon of the highly
fossiliferous (but non-

§°

~£ c ^

soni
28. Cyrtograptus symmetri-

graptolitic) Wenlock
limestone of Wenlock

ri 0)
0 >

u£

«

cus

27. Monograptus ric car-
ton en sis *
26. Cyrtograptus murchi-

Edge (Shropshire),
Dudley, Walsall, etc.
For Graptolites,
Builth (Radnor), Ric-

SOHl *

k carton (S. Scotland)

25. Monograptus crenulatus

d

rt

24. Monograptus griesto-

(_i
rt

ncnsis

Galashiels (S. Scot-

H a

23. Monograptus crispus

land), Tarannon

»* Q •

o a

22. Monograptus turricu-

River (Central

J3

latus

Wales)

13

Band of Rastrites

O

maxim us

'

VALENTIAN '

2 1 . Monograptus sedgwicki *
Band of Cephalograp-
tus cometa *

For Graptolites,

.20. Monograptus convo

Dobb's Linn (Dum-

_

lutus *

friesshire), Dona-

5

19. Monogyaptusgregarius :

ghadee (Ireland)

^4 <

M. argenteus^

>For Brachiopods, etc.,

M. triangulatus ^

May Hill (Glouces-

M. fimbriatus f

tershire), Malvern

1 8. Monograptus cyphi'S*

(S. Shropshire), Gir-

17. Mesograptus modestus

van (Ayrshire)

and Orthograptus

vesiculosus

'

396

PALEONTOLOGY

11. ORDOVICIAN GRAPTOLITE ZONES.

After Lap worth, Elles and Wood.

Series or Stages.

Graptolite Zones.

Famous Localities.

ASHGI LL IAN

(Upper Hart--
fell) |

16. Ceplialograptus acu-
minatus *
15. Dicellograptus anceps
14. Dicellograptus com-
planatus *

IDobb's Linn, near
Moffat (Dumfries-
shire)

CARADOC i AN
(Lower Hart-
fell)

13. Pleurograptus line-
arts *
12. Dicranograptus din-
gani *
1 1 . Climacograptus wil-
soni

Dobb's Linn and
Hartfell (Dumfries-
shire)

LLANDEILIAN
(Glenkiln) '

10. Climacograptus pelti-
fer and Mesograp-
tus multidens
9. Nemagraptits gra-
cilis * f
8. Glyptograptus tere-
tiusculus

Glenkiln Burn (S.
Scotland)

LLANVIRNIAN-

7. Didymograptits nnir-
chisoni
6 Didymograptus Ufi-
rfttSf

[Abereiddy Bay (Pem-
I broke)

SKI DD AVIAN
(Arenig)

5. Didymograptus hi-
rundo
4. Didymograptus ex-
tensus
3. Dichcgraptus

(Skiddaw (Cumber-
land)

TREMADOCIAN i

2. Bryograptus *
i. Die tyonema sociale *

Pedwardine (Hereford-
shire)

NOTE. — Zones marked * occur in the same sequence in Scandinavia as in Britain,
and those marked f in North America. The identity of other zones may be masked
by difference of name; or greater or lesser thickness of sediment or abundance of
fossils may lead to a larger or smaller number of zones being recognized ; or zones
may be wanting through non-sequence.

DIVISIONS OF GEOLOGICAL TIME

397

12. CAMBRIAN TRILOBITE ZONES.

After Walcott, Moberg, Fearnsides, Illing, and others.

Series.

Zones.

Famous Localities.

A ngflina sedgwicki

Tremadoc (N. Wales)

TREMADOCIAN
(Transition to -
Ordovician)

Bryograptits and Shu-
mar di a
Asaphellus homfrayi
Dictyoncma sociale

I Shineton Brook
f (Shropshire)

Pedwardine (Hereford-

shire)

Niobe homfrayi

Penmorfa (N. Wales)

Peltura scarabaoides

Scania

Spharophthalmus alatus

White-leaved Oak

(Malvern)

POTSDAMIAN ...«

Eurycare latum
Orthis lenticularis

Portmadoc district (N.

-Wales)

Olenus truncatus

Hartshill and Oldbury

(Warwickshire)

Paradoxides fo rchammeri
P. davidis

St. David's (Pembroke)

IConocoryphe

Bornholm (Baltic Sea)

aqualis

MENEVIAN ...-
(Acadian)

Agnostus par-
vifrons

Scania

Ctenocephalus

Scania

exsulans

P. olandicus

—

Protolcnus

Comley (Shropshire),

Newfoundland

Ohndlus
Callavia

Comley (Shropshire)

Elliptocephala

—

- Nevadia.

—

APPENDIX II
STRATIGRAPHICAL PALEONTOLOGY.

I.— GENERAL FEATURES OF

OLDER PALAEOZOIC FAUNAS TAKEN

AS A WHOLE.

PRESENCE in abundance of Graptolites., Brachiopods,
and Trilobites ; in fair abundance of CORALS, CRINOIDS,
and CYSTIDS, NAUTILOIDEA (mainly straight or slightly
coiled). Rarity of Lamellibranchs and Gastropods.
Complete or almost complete absence of Echinids,
Ammonoidea, Vertebrata, and Plants.

SPECIAL FAUNAL FEATURES OF THE SEPARATE
SYSTEMS.

1. CAMBRIAN.— Absence of Graptolites (except
in uppermost beds), Corals, Crinoids, and Cephalopods.

BRACHIOPODS are mainly horny Inarticulata ; the
Orthids are the only other common forms. TRILOBITES
are micropygous (except Agnostus) ; they serve as zone-
fossils. The index-species of the Upper Cambrian are
mainly of the family Olenidtz, those of the Middle Cam-
brian, Paradoxidce, while Walcott's rather tentative
divisions of the Lower Cambrian are based on genera of
Mesonacidce. For details see p. 397.

2. ORDOVICIAN.— Abundance of branched Grap-
tolites, the most complexly branched being the earliest.
The Graptolites serve as zone-fossils (see p. 396), but are
almost confined to the black-shale facies. The Lower
Ordovician (Skiddavian-Llanvirnian) is characterized by
Dichograptids, the Upper by Leptograptids (Llandeilian),
Dicranograptids and Diplograptids.

399

400 PAL/EONTOLOGY

BRACHIOPODS : Orthids and Strophomenids abun-
dant ; horny Inarticulata reduced in numbers ; Rhyn-
chonellids and Spire-bearers appear, but are rare.

TRILOBITES of nearly all families abundant ; most
distinctive are the Trinucleids and Asaphids. Trilo-
bites may serve as zone-fossils in beds of " shelly " facies.

3. SILURIAN.— Few branched Graptolites; Mono-
graptus is the predominant genus, and its species
(chiefly) serve as zone-fossils (p. 395).

Corals and Crinoids very abundant in the limestones.

BRACHIOPODS : Pentamerids first appear and are
abundant ; Ordovician forms continue ; Rhynchonellids
and spire-bearers (Atrypa, Mevistina, Spirifer) common.

TRILOBITES as in Ordovician except for absence of
Trinucleids and Asaphids. Encrinurids and Illsenids do
not survive beyond this system.

VERTEBRATA (primitive Fishes, represented by skin-
teeth and fin-spines) first appear near the top of the
system.

II.— GENERAL FEATURES OF

NEWER PALEOZOIC FAUNAS TAKEN

AS A WHOLE.

Fishes and land-plants appear and become abundant.

Corals, Crinoids, and Brachiopods continue abun-
dant. Among the latter, Loop-bearers (Tevebmtnlids)
first appear, and spine-bearers are very abundant.

ECHINOIDS and BLASTOIDS, previously very rare,
become fairly common, especially Blastoids, which,
however, die out in this era.

LAMELLIBRANCHS and GASTROPODS increase in
numbers.

AMMONOIDEA (Goniatites) first appear and become
abundant.

NAUTILOIDEA continue common, and fully-coiled forms
become more frequent.

TRILOBITES and CYSTIDS become rarer.

Graptolites are completely extinct.

STRATIGRAPHICAL PAL/KONTOLOGY 401

SPECIAL FEATURES OF THE SEPARATE SYSTEMS.

4. DEVONIAN. — BRACHIOPODS: of horny Inarticu-
lata, only Lingula and Discinids (Orbiculoidca] survive
(lasting to the Recent period). Terebratuloids first appear,
but are not common. Most other Silurian families
survive. Spire-bearers abound.

Goniatites are common, with very simple suture-
lines ; they serve as zone-fossils.

TRILOBITES : much reduced in numbers, but the same
families as in the Silurian except Illanids and Encrinurids.

FISHES very abundant, especially Ostracoderms and
Ganoids.

LAND-PLANTS (PTERIDOSPERMS, etc.) not very common.

5. CARBONIFEROUS. — Blastoids commoner
than in any other system.

BRACHIOPODS : ProductllS, Spirifer, and Rhynchonel-
lids most abundant. Terebratulids fairly abundant.

CEPHALOPODA : Goniatites with suture-lines a little
more complex than in Devonian ; tightly-coiled NAUTI-
LOIDS.

TRILOBITES rare, of only one family (Proetida).

VERTEBRATA: fish- teeth and spines common in some
beds. OSTRACODERMS are extinct. AMPHIBIA in Coal
Measures, rare.

PLANTS abundant in the coal-bearing beds. In the
northern hemisphere (except India) the fern-like leaves
of Pteridosperms, and stems of Lepidodendron,
Sigillaria and Calamites are common. In India and
the southern hemisphere, there is a totally distinct flora
(Permo-Carboniferous), with the fern-like Glossopteris.

The zoning of the Lower Carboniferous is based on
shallow- water Corals and Brachiopods, and the zones may
be of restricted geographical extent (pp. 392-3). The
Upper Carboniferous (Coal Measures) are classified, but
scarcely zoned, by the flora : their marine equivalents by
Goniatites, as are the marine Permian.

6. PERMIAN. — The true marine fauna is only
known in a few restricted (tropical and sub-tropical)

26

402 PALAEONTOLOGY

regions. It is very similar to the Carboniferous fauna,
except that the Ammonoidea have far more complex
suture-lines, frequently with many-pointed lobes and
rounded saddles.

In other areas an impoverished (inland sea) fauna is
found, consisting mainly of BRACHIOPODS and LAMELLI-
BRANCHS, and a terrestrial flora which is an impoverished
form of the Carboniferous flora.

III.— GENERAL FEATURES OF MESOZOIC
FAUNAS.

CORALS are less abundant and of entirely new types.

Cystids and Blastoids are extinct, CRINOIDS become
less and less common ; Echinids are abundant, and incliif'-
Irregular as well as Regular forms.

Brachiopods are common, especially Terebratulids
and Rhynchonellids.

Lamellibranchs are abundant, and Gastropods are
common.

Ammonoidea are very abundant, except towards the
end of the era ; their sature-lines are extremely complex
as a rule. They serve as zone-fossils (pp. 384-390).

Trilobites are quite extinct.

VERTEBRATA: Fishes (sharks and Ganoids) and
Reptiles, both terrestrial and marine, are abundant.
MAMMALIA are exceedingly rare and small.

LAND-PLANTS : principally Cycads, also CONIFERS and
FERNS.

SPECIAL FEATURES OF THE SEPARATE SYSTEMS.

7. TRIASSIC-— As with the Permian, though to a
much less extent, the true marine fauna is restricted in
area, being preserved principally in the Alps and other
great mountain chains.

ECHINOIDS are rare and of very special types.

Brachiopoda : spire-bearing forms are still numerous,
as well as Terebratulids and Rhynchonellids, but nearly all
other Palaeozoic families are extinct,

STRATIGRAPHICAL PALEONTOLOGY 403

CEPHALOPODA : Ammonoidea with ceratitic suture-
lines are very abundant, but there are others with more
complex (ammonitic) suture-lines.

Primitive BELEMNOIDEA also occur.

VERTEBRATA : Reptiles, especially terrestrial forms,
are abundant. Some marine reptiles.

ZONE-FOSSILS : Ammonoids (and Lamellibranchs).

8. JURASSIC. — Echinoids are abundant, especially
in certain beds. Clypeus is confined to this system,
Hemicidavis nearly so. Acrosalenia, Holectypus, Pygaster,
Nucleolites, and Collyrites are Cretaceous also.

The last spin-bearing Brachiopods occur in the Lower
Jurassic.

Lamellibranchs are abundant, especially Trigonia
and Oysters.

Ammonoidea with highly complex suture - lines
(Ammonites) and typical Belemnites are abundant.

Large marine and terrestrial Reptiles are common.

LAND-PLANTS (CYCADS, etc.) are common in estuarine
deposits.

The Lower Jurassic is characterized by the presence
of the brachiopod Spiriferina, the absence of Irregular
Echinoids, and the dominance of keeled forms among
the Ammonites (though in particular zones this dominance
may be reversed). The Upper Jurassic is characterized
by the absence of Spiriferina, the presence of Irregular
Echinoids, the dominance of unkeeled Ammonites
(reversed in particular zones), and the frequency of
lappets on the apertural margin of Ammonites (whether
keeled or unkeeled).

ZONE-FOSSILS. — Ammonites or, in their absence,
Brachiopods (pp. 385-9).

9. CRETACEOUS.— Echinoids of several new
families appear, among which the Echinocorythidce and
Spatangidte at once become abundant.

In the Alpine-Mediterranean province, numerous very
peculiarly-shaped inequivalve Lamellibranchs (Rudistes :
Monopleura, Caprotina, etc.) occur in the Lower Cretaceous,
and are followed in the Upper by the Hippurites, whose
habit of growth is more like that of rugose Corals than
Lamellibranchs.

404 PAL/EONTOLOGY

The last British Trigonia is found in the Cenomanian
(Middle Cretaceous.)

Ammonoidea and Belemnoidea become gradually
less abundant and finally die out. More or less uncoiled
Ammonoids are common.

In certain regions (especially North Africa, Syria,
South America, and the southern U.S.A.) there are
Upper Cretaceous Ammonoids with suture-lines closely
resembling those of the Triassic Ceratites (pseudo-
ceratites).

In the Upper Cretaceous there is a sudden appearance
of Fishes and land-plants (Dicotyledons) of much
more modern types than those of the Jurassic and Lower
Cretaceous.

ZONE-FOSSILS. — Ammonites and Belemnites, supple-
mented by Brachiopods and Echinoids in Northern
Europe, by Rudistes and Foraminifera in the Mediter-
ranean province.

IV.— GENERAL FEATURES OF THE
CAINOZOIC FAUNAS.

Foraminifera are abundant and include forms of
much greater size than in most earlier systems. These,
where they occur (mainly in tropical and sub-tropical
regions), are the best zone-fossils.

Corals are not abundant except in warm latitudes.

Crinoids and Brachiopods are rare.

Echinoids are common, especially in warm latitudes,
where two new families abound : the flat Scutellids and
the flat or pyramidal Clypeastvids. Spatangids increase in
abundance.

Lamellibranchs and Gastropods are extremely
abundant, and include many new families. Their
species have too long a range to serve as zone-fossils,
but the age of a bed can often be fixed exactly by the
association of several species, some of which range
upwards and others downwards.

CEPHALOPODA are rare, Ammonoids and typical
Belemnoids being extinct.

STRATIGRAPHICAL PALAEONTOLOGY 405

VERTEBRATA: Fishes are of modern types; many
Mesozoic types of Reptiles are extinct, only those
surviving which last to the Recent period (crocodiles,
turtles, lizards, snakes). Mammals are abundant,
mainly in freshwater deposits, and serve to some extent
as zone-fossils.

PLANTS are of modern types, but many now confined
to warm climates (such as palms) are found fossil as far
north as the Arctic regions.

SPECIAL FEATURES OF THE SEPARATE SYSTEMS.

10. EOCENE.— The foraminifer Nummulites is
very abundant, especially in the Middle Eocene. Ortho-
phragrnina is confined to this system, except in America.

The Echinoids Conoclypeus and Nncleolites survive frofn
the Cretaceous, but not beyond the Eocene.

The same is the case with the Pycnodonts (GANOID
fishes).

The Molluscan fauna of this period (in Britain and
elsewhere) resembles that now living in the Indian
Ocean.

The Mammalia of the Lower Eocene are all small
and primitive, with forty-four low-crowned teeth and
five-toed limbs. In the Middle and Upper Eocene they
rapidly become larger and more specialized.

TI. OLIGOCENE.— The foraminifera Nummulites
and allied genera continue in some abundance. Lepido
cyclina replaces Ovthophvaginina.

The Irregular Echinoids Scntella and Clypeaster first
appear.

Mammals. — In Europe and America new families
appear, including the first (hornless) rhinoceros. In
Africa a distinct fauna was developing, including the early
ancestors of the elephant.

12. MIOCENE. — Nummulites is practically extinct.
Lepidocyclina persists at first, but is finally replaced by
Miogypsina.

The Spatangids Micraster and Holaster, which have

460 PALEONTOLOGY

survived from the Cretaceous, now appear for the last
time, alongside Schizaster and others which continue to
the Recent period.

Mammalia include the first antlered deer, first horned
rhinoceros, first ape. Ancestors of elephants suddenly
appear in Europe.

13. PLIOCENE. — During this period there is in the
northern hemisphere a gradual southward migration of
the marine fauna, Arctic species appearing in Britain and,
later, in the Mediterranean area. The horizon of any
normal marine bed in the Pliocene system can be deter-
mined by the percentage of its species which are
(a) extinct, (b) surviving in more northern or (c) in more
southern latitudes.

Mammalia attain their greatest size. The first one-toed
horse appears. Mastodon is characteristic.

INDEX

ABOKAL poles, 261

Anaptychus, 161, 172

Astrauida;, 311

Acanthoceratidae, 179

Anarccstcs, 170

Atelostomata. 276

Acanthodii, 236

Anatinacea, 91

At/nris, 42, 43

Acanthothyris, 37

Ancyloceras, 144, 173, 178

Atiactites, 186

Accretion, Growth by, ir,

Angiosperms, 340, 342

Atremata, 22, 23

12, 263
Acervularia, 310

Angle of Divergence, 284
Anisomyarian, 65

Atry/>a, 31, 42
Aturia, 750

Acidaspidae, 210

Annelida, 329

Aulacoceras, 187

Acme, 19

Annularia, 338

.rf ulacotkyris, 4 1

Acrosalenia, 273

Anomaljna, 323

Art lop ra, 314

Acrofreta, 25

Anomia, 88

Aurinia, 117

Actaon, 119

Anomodontia, 237

Aves, 237

Actinal Furrows, 276

Antedon, 249, 255

Aviculopecten, 89

Actinoccunax, 187

Antennas, 203, 218

Axial Canals, 247

Actinoceras, 149

Anterior Canal, 101

Axonolipa, 290

Actinocrinus, 254

Apical Disc, 264

Axonophora, 290

Actinocionta, 77

Apiocrinus, 255

Aymestry Limestone, 33

Actinopteryeii, 228

Apodeme, 195

Actinozoa, 797, 309

Aporita, 258

Bactrites, 170

Adacna, 80

Aporosa, 311

P.aculicone, 143

Adaptation, 73, 76, 205, 231

Aporrkais, 115

Bacitlites, 142, 169, 179

Adductors, 10, 13, 26, 50

Appendages, 201

Barrande, J., 189, 214, 279,

Adesmacea, 92

Aptychus, 160, 172

S73

.•Jtglina., 209

Apus, 202

Barroisia. 319

Agnostits, -210, 211

Arachnida, 215, 219

Barton Beds, 47, 61, 106

Agfftuatifts, 170

Araucciria, 342

Basals, 244

Alar septa, 301

Area, 85

Bather, F. A., 250, 279

Alcyonaria, 313

Arcestes, 171

Beecher, C. E., 21, 206, 214,

Alectryonici, 89

ArcJueocalamites, 338

219 .

Algae, 333, 347

Archceocidaris, 272

Beekite, 298, 346

Alethofiteris, 340

Archceoniscus, 217

Beleinnitella, 186

Atveolinn, 325, 327

A rchieopteryx, 233

Belemnites, 182

Alveolites, 314

A rchccozonites, 1 22

Beleutnopsis, 186 •

Alveolus, 183

Archanodoii, 77

Beleinnoteuthis, 18^

Amaltheidae, 161, 174, 175

Arenacea, 322

Bellero/>hina, \\)

Ambitus, 261

Argonauta, 124

Bellerophon, in

Ambulacra, 260, 262

Arietidae, 161, 173

Belosepia, 187

Ambulacrals, 246

Aristocystis, 258

Bembridge Limestone, 107

Ammonites, 169

Arnioceras, 173

Benthic, 145, 198, 213, 218,

Ammonoidea, 148, 384-390

Arthrodira, 236

226

Amphibia, 229, 237

Arthropoda, 191, 192, 215

Bernard, F., 93, 372

Amphicoelous, 228

Articulata, 22, 250

Biloculina, 321, 325

Amphidetic, 51

Artiodactyla, 235, 238

Binominal System, 361

Aitiphoracrinus, 247, 254

AsaphidcK, 207, 212

Biramous, 202

Amphoridea, 258
Aniplexus, 303, 310

Ascoceras, 150^
Aspidobranchia' no

Krds, 233, 237
Biserial, 273, 288

Ainpy.v, 209, 212

Aspidoceratidae, 177

Bivium, 277

Anabacia, 313

Assilina, 324

Blastoidea, 243, 259

Anagenesis, 136, 143, 167,

Astarte, 84 ^

Body-Chamber, 125, 147

308

Asteroceras, 129, 173

Bog nor Rock, 49

Anal Plate and Tube, 248

AsteropkyUites, 338

Borings, 318, 347

Anal Pyramid, 256

Astogeny, 287

Boss, 265

407

4o8

PALAEONTOLOGY

Bothriocidaris, 271

Caudal Fins, 228

Cordaites, 341

Bourgueticrinus, 255

Caudalization, 205

Cornbrash, i, 17, 64

Bourrelet, 277

Cephalaspis, 222

Corona, 262

Brachial Loop, 5, 6, 9, 12,

Cephalization, 205

Coronate, 165, 174, 176, 17

39, 40

Cephalon, 191

Co>moceratida;, 177

Brachials, 246

Cephalopoda, 123

Costa, 147, 307

Brachyodont, 234
Brachyura, 21.5, 218
Bradford Clay, 18, 255

Cementation, 31, 32, 46, 66
Centrechinoida, 273
Crntronella, 38

Costate, 167
Cotype, 369
Counter-Septum, 301

Ceraiites, 169, 171

Crag, 105, 329

B*yog>af>t*s< 286

Ceratitic Suture, 163

Crania, 25

Bryophyta, 331
Bryozoa, 21, 287, 328
Buccinum, 116

Ceratodus, 229, 230
Ceritliium, 101, 113
Cetacea, 238

Cranidium, 197
Crassatella, 61, 84
Crenulate, 50, 265

Buch, L. von, 169

Chama, 82

Crepidula, 112

Buckmla, 77

Chara, 334

Crinoidea. 243

Buckman, S. S., 44, 139.

Ckeiru*us, 210

Crioceras, 173, 178

165, 172, 189, 190, 381,

Chelonia, 233, 237

Criocone, 142, 178

Chert, 317, 320

Cristellaria, 322

Bulla i IQ

Chilidium, 26, 27

Crocodilia, 237

Burial, 48, 49» H6
Burlingia, 210
Burrowing, 347
Byssus, 65

Chlamys, 88
Chondrichthyes, 222, 236
Chonetes, 31, 348
Cidaris, 272

Crossing Canal, 284
Crossopterygii, 228
Crotalocrinus, 250, 253
Crown, 225, 243, 244

Cirripedia, 215, 217

Crura, 14

Cadicone, 165, 177

Cis ell a, 40

Crural Plates, 34

Cladoselachii, 236

Crustacea, 215

Cadornella, 43

Classification, 21, 72, 204

Cryptocrinus, 258

Catamites, 338, 342
Calamoclculus, 338
C alamos achys, 338

Clavella, 1 1 7
Climacograptus, 288, 293
Cliona, 318, 347

Cryptogenetic, 181, 235, 34:
Ctenobranchia, 112
Ctenodonta, 76, 77

Calcaire Grossier, 101, 327,

Clisiop'iyllum, 310

Cuculltea, 86

334

Cliiaitibonites, 35

Cupressocrinus, 243, 250

Calcarina, 323

Clono^raptus, 2J6

CyatJui.ronia, 310

Clymenia, 394

Cvat/iocrinns, 253

Calcispongiai, 319
Callavia, 211
Callus, too

Clypea..itert 276
Clyf-eu*, 277
Coal Measures, 121, 219

Cyathophyllum, 310
Cycadeoidea, 341
Cycadophyta, 340

Cochliodontidae, 225

Cycle, Evolutionary, 143

Calyptru-a, 113
Calyx, 244, 298
CamarophoriOi 34, 35
Cainarotcechia, 36
Camerata, 248, 250
Caninia, 303, 310
Capricorns, 174
Capulus, 112

Ccelenterata, 297
Coeloceras 174
Cwloptychiuni, 319
Ccrlotetithis, 186
Collyrites, . 77
Columella, 99
Columnals, 243, 304
Common Canal, 283

Cyclolites, 313
Cvclolobus, 171
Cyclostomata, 236
CyclotJiyris, 37
Cylindroteutkis, 186
Cyphosoma, 273
Cyprina, 84
Cy/>r<ea, 106

Conchidium, 31, 33

Cyrcna, 57, 84

Carboniferous Limestone,

Conchiolin, 53

Cyrenoid Hinge, 59, 60

171, 247, 253, 254, 298,
704 310 327) 393

Congeria, 90
Coniferales, 342

Cyrtia, 31, 42

( 'ytina, 42

Cardinal Area, 4, 47

Conocardium, 86

Cyrtocone, 139, 149

Cardinal Process, 9, 10
Cardinal Septum, 300
Cardinal Teeth, 58

Conocoryphida, 207
Conorbis, 117
Conoteuthis, 189

Cyrtografi'.us, 295
Cystiaea, 243, 255
Cystipkyllun', 311

Cardinia., 79

Conothecal Stride, 184

Cardiocarpus, 341

Constrictions, 126, 176

Dacryomycii 76
Uactyliicer&s 175

Cardioceras, 177
C avdiola, 77

Conulus, 265, 275

Dactyloids, 147, 175

Cardita, 84

Convs, 117

Dactylotcutliis, 187

Cardiunt, 80
Carina, 147, i63, 266
Cassiduloidea. 277

Convergence, 73, 222, 232
Cophinoceras, 138
Coral Rag, 260

Dallina, 39, 40
Dahnanella, 30, 32
Dalmanites, 198, 211

Casts, 23, 146, 346
Catagenesis, 1331 136, 143,
163, 167, 308

Corbicula, 61, 84
Cor bis, 82
Corbula, 69, 92

Davidsonia, 28
Decapoda, 215, 218
Deiphon, 211

INDEX

409

Deltidial Plates, 4, 13

Duvalia, 187

Filicales, 335, 339

Delthyris, 42

Dysodonta, 86

Fins, 226, 228, 230-232

Delthyrium, 4, 13

Fishes, 222

Deltoids, 246

Ecdysis, 192

Fissuridea, in

Demi-Plates, 263

Ech.inoca.ns, 217

Flabellum, 311

DendrocystiS) 258
Dendrophyllia, 313

Echinocorys, 278
Echinocyamus, 276

Flares, 173
Flexibilia, 250

Dendropupa, 122

Echinocystis, 271

Floscelle, 277

Dentalium, 329

Echinoidea, 243, 347

Food-Grooves, 246

Dental Plates, 14

EchinolanipaS) 277

Footprints, 347

Dental Sockets, 9

Echinosf>h<zra, 255, 258

Foramen, 4

Dentine, 225

Echinothuria.) 273

Foraminifera, 319, 320, 382

Derived Fossils, 69, 183,

Echinus, 273

383- 392

225, 348, 374

Echioceras, 174

Forest Marble, 17, 18

Derbya, 28, 30

Edestidae, 225

Fossula, 298

Deroceratidae, 174

Edtnondia, 91

Freeh, F., 290, 372, 391

Desmoceratidae, 178

Edrioasteroidea, 243

Freshwater, 218, 227

Desmodonta, 91

Elasmobranchii, 236

Fungida, 313

Dextral Shells, 98

Elastic Ligament. 47

Fusiform, 103

Diatomaceae, 333

Eleutherozoa, 242, 243 260

Fusulina, 324, 327, 392

Dibranchiata, 147, 149, 182

Elles, G. L., 290, 294, 314,

Fusus, 1 06, 117

Dibunophyllum, 310

395, 396

Dicellograptus, 291

Elfaptocephala, 211

Gametophyte, 331, 332

Diceras, 81

Etnarginala^ 95, in

Gammi-Radiate, 168

Dichograptidae, 287, 290

Enamel, 225

GangatKopteris, 340, 343

Dicranogr.tpius, 291

Encrinuridse, 210. 212

Ganoidei, 227

Dictyontma, 295

Encrinus, 254

Gas-Chambers, 125

Dicyclic, 244, 250

Endoceras, 149

Gastrioceras, 154, 171

Didytnograptus, 282
Dielasma, 38

Endocyclica, 271
Endogastric, 137

Gastropoda, 94
Gemmellaro, G. G., 170

Dimorphism, 321

Endopodite, 202

Gena, 194

Dimorpliograptus, 29 j

Endothyra, 326

Genital Plates, 264

Dimyarian, 52

Ensis, 82

Genotype, 370

Dinosauria, 237

Entomostraca, 215

Genus, 22, 355, 356

Diplodocus, 237

Entosepta, 299

Geographical Distribution,

Diplodo-ita, 82

Epitheca, 311

7, 53, 62, 107, in, 115.

Diplografitus, 293

Epithyris, 21, 39

145, 172, 179, 213, 227,

Difilopodia, 273

Equisetales, 335, 337

229, 238, 296, 341,342, 343

Diplopora, 334

Equisrtites, 338

Geological Range, 211, 212

Diploporita, 259

Equivalve, 46

Gephyroceras, 171

Dipnoi, 22-}, 236

Equus, 235

Gerontic, 79, 88

Discina, 16, 25

Erodona, 69

Gervillia, 86

Discinisca, 15, 16, 25, 30

Escutcheon, 52

Gills, 198, 202

Discoidal, 165

Estheria, 216

Ginkgo, 341

Discoidea, 275

Etheridgina, 31

Gissocnnus, 253

Discorbina, 323

EuoMiphalus, in

Glabella, 194

Dissepiments, 302

Eucalyptocrinus, 254

Glandina, 121

Distortrix, 105

Euechinoidea, 271

Glauconite, 327, 347

Divaricators, 10, 13, 26
Docoglossa, no

Eufilicinae, 339
Euiypterus, 215, 218

Globigerina, 323, 327
Glossograptidae, 294

Dollo, L., 200, 219

Euthyneura, 109

Glossopteris, 340, 343

Dona r, 84

Evolution, 269

Glossotfiyris, 39

D'Orbigny, A., 169, 189,

279' 373
Dorsal Cup, 244
Doryderma, 318

Evolutionary Cycle, 143
Excavations, 288
Exhalent Aperture, 51
Exocyclic, 273

Glvphioceras, 171
Glyptosphccra., 259
Gnathobase, 202
Gnathostomata, 275

Dosinia, 85

Exogastric. 137

Goldfuss, G. A., 374

Djuville. H., 76, 93, 189,

Exogyra, 8;

Gomphoceras, 138, 150

330, 382

Exopodite, 202

Goniatites, 157, 169, 394

Douville, R., 330, 382

Exosepta, 299

Goniopkyllum, 311

Drcisscnsitt oo

Goniopora, 313

Duerden, J. E., 301, 315

Facial Suture, 196

Grammysia, 91

Dumortieria, 175

Family, 22. 355, 368

Graptolites, 281, 3<:o, 395

Duncan, P.M., 271, 279,

Fasciole, 268, 278

396

315

Favo ites, 313

Gryphaa, 89

Duraniii, 82

Fibularia, 276

Guard, 183

4io

PALEONTOLOGY

Guide-Line, 134

Hydroid, 296

Lantern of Aristotle, 264

Gymnosperms, 340
Gypidula, 34

Hydropore, 256
Hydropteridae, 339

Lappets, 153, 160, 168, 176
Lapworth, C.,29o, 296, 314,

Gyroceras, 170

Hydrospire, 260

395, 396

Gyrocone, 139, 170

Hydrotheca, 283

Larviformia, 250

Gyroporella, 334

Hydrozoa, 281, 297

Lateral Area, 132

Hyinenocaris, 217

Lateral Teeth, 58

Habitat 18
Haliotis, in

Hy tithes, 120, 121
Hyponomic Sinus, 128

Lectot) pe, 369
Leda, 76

Hall, J., 44, 374

Hypoparia, 206

Lepadocrinns, 258

H a 'y sites, 314

Hypostome, 197

Leperditia, 217

Hamites 144, 169, 179

Hypothyrid Foramen, 35

Lepidocyclina, 325, 327,

Hammatoceratidae, 175

Hypsodont, 235

328, 382

Harpedidae, 209
Harpoceras, 175

Ichthyodorulites, 223

Lepidodendron, 338, 342
Lepidostrobus, 338

Hastate, 184

Ichthyosaurus, 228, 231,

Lepidotus, 227

Hastites, 187

233

1. eptcena, 26

H alter la, 231, 237

Ichthyotomi, 236

Leptograptus , 291

Helicoprion, 226

Iguanodon, 237

Lichadidae, 210

He Halites, 314

Illcznus, 207, 212

Ligament, 47

Heliophyllum, 310

I mperforata, 322, 326

Lima, 89

Helix, 121

Imperforate, 99

Limbs, 198, 201

Hemera, 19, 20, 381

Impressed Area, 132

Limncea, 121

Hemiaster, 278

Impressions, 335

Liinnocardiuni, So

Hetnicidaris, 260, 273

Inaduriata, 250

Lingitla, 16, 21, 30

Hemithyris, 13

Inarticulata, 22

Lingidella, 24

Hercoglossa, 150

Indexes, 371

Linnaean System, 356, 361

Herpetocrinus, 251

Inequilateral, 56

Lioceras, 175

Heterocercal, 226

Inequivalve, 47

Liothyrina, 39

Heterodonta, 57, 80

Inexpleta, 309

Liparoceratidae, 174

Heteroniyarian, 65

Infra-Basals, 244

Lithistida, 318

Heteropoda, 119

Inhalent Aperture, 51

Lithophaffus, 90, 348

Heteropygous, 205

Inoceramus, 86

Lithostrotion, 304, 310

Heterosporous, 334

Insecta, 216, 219, 345

Lithothatnnion, 334

Heterostrophic, 109, 112

Integripalliate, 52

Littorina, 112

Hexacoralla, 302

Interambs, 262

Lituites, i;o, 293

Hexactinellida, 318

Interbrachials, 248

Llandovery Beds, 309

Hexapoda, 216 •

Interporiferous Areas, 270

Lobe, 134, i6i| 163

Hibolites, 186

Interradii, 245

Loculus, 302

Hildoceras, 175

Intussusception, 263

Loganograptus, 286

Hinge, 9, 10, 50, 52, 57-60
Hippopodium, 79
Hippurites, 81

Irregularia, 270, 273
Irreversibility, 143
Isastrcpa, 311

Loligo, 189
London Clay, 218
Lonsdaleia, 310

Holarctic, 238

Isnienia, 40

Loop, 5, 6, 9

Holaster, 278

Isocardia, 84

Ludna, 8?

Holectypus, 275
Holm, G.. 282, 314

Isocrinus, 254
Isomyarian, 52

Lucinoid Hinge, 59, 60
Ludlow Beds, 136, 222, 223

Holocystis, 311

Isopoda, 215, 217

JLudiuigia, 175

Holoptychiits, 230

Isopygous, 205

Lunule, 61, 276

Holostome, 99, 102, 112

Lycopodiales, 334, 338

HolotHuroidea, 243, 278

Jugum, 41

Lyginodendron, 340

Holotype, 369

Lyra, 41

Homalonotus, 209

Keel, 132, 147, 168, 172

Lyrodestna, 77

Homocercal, 227

Kepplerites, 177

Lytoceras, 172, 178, 181

Homoeomorphs, 28, 29, 180

Koninckella, 43

H ' omaeorhynchia, 37

Kutorgina, 25

Macrocystella, 258

Homonym, 361

Macroscaphites, 144, 173,

Homosporous, 334

Labrum, 197, 267

178

Hooke, Robert, 17

Lagena, 322

IMactra, 82

Hoplitidas, 179

Lagenostoma, 340

Macrura, 215, 218

Horizontal, 284

Lamarck. J. I». de M., 374

Madreporaria, 309

Huronia, 149

Lamellibranchia, 45, 347

Madreporite, 264

Hyalina, 322

l.anarkici, 223

Magas, 40

Hyatt, A. S., 169, 189, 190

Lanceolate. 184

Magellania, 7, 39, 40

Hyboclyfeus, 273
Hydrocorallines, 296

Lancet- Plates, 259
Lang, W. D., 308, 315, 381

Magnosellaridae, 170
Malacostraca, 215

INDEX

411

Mamelon, 265

Myriopoda, 215

Ornithosauria, 237

Mammalia, 234, 238

Mytilus, 90

Orontetopits, 204, 209

Mammoth, 345

Orthis, 32

Manticoceras, 151

Nacreous, 53

Orthoceras, 124

Marattiales, 339

Naiadacea, 77

Orthocone, 139, 149, 170

Marginopora, 325

Nassa, 116

Orthonota, 91

Marsipocrinus, 253

Nassellaria, 3 o

Orthophragtnina, 324, 327,

Marsupialia, 234, 238

Natica, 99

382

MarsupiteS) 248

Nautilinidae, 170

Orthotetes, 28, 31

Martinia, 42

Nautiloidea, 148, 149

Osmunda, 339

Medlicottia. 171

Nautilus, 124, 126, 150

Osteichthyes, 227, 236

Mcdullosa, 340

Nectic, 145, 199, 218, 226

Ostracea, 89

Meekella, 28, 31

Nema, 282

Ostracoda, 215, 217

Meekoceratidae, 171

Nctnagraptus, 291

Ostracodermi, 222, 236

Megalodon, 79

Neotremata, 22, 24

O street, 66, 90

Megalospheric, 321

Neptunea, 116

Otozamites, 341

Megaspore, 332

N erinceat 115

Ovulites, 334

M egateut/iis, 187

Nerita, 112

Oxycone, 165

Megathyris, 39, 40

Neritinct) '112

Oxynoticeras, 173

Megerlina, 40

Neumayr, M.,74, 169, 181,

Uxyteuthis, 186

Melanin., 113

190, 372

Melanopsis, 113

Neuropteris, 340

Pachyceratidae, 177

Meleagrina, 66

Nevadia, 211

Pachypora, 314

Mellita, 276

Nilssonia.) 341

Pachyrisma, 79

Meloceras, 138

Nipponites, 144

Pachyteuthis, 186

Melonechinus, 271

Nodosaria, 322

Pagetia, 210

Meretrix, 63, 85

Normal Line, 134

PeUaecnitats, 271

Meristina, 43

Notch, Marginal, 96

Palteocorystes, 218

Merostoinatx, 215

Nncleolites, 277

Pa/ceocyclus, 309

Mesonacis, 207, 211

Nucula, 52, 76

Palceodisciis, 271

Metasepta, 301

NummuliteS) 324, 327,

Palceospondylus, 236

Metatype, 370

328. 382

Palechinoiaea, 271

Michetinia, 314

Palingenesis, n, 12, 13

Micrasier, 266, 278

Obolus, 24, 30

Pallial Line, and Sinus,

Microderoceras, 174
Microdiscus, 210

Obverse, 284
Occipital Furrow, 194

Palpebral Lobe, 197

Microphagous, 247

Ocular Plates, 264

Panopea, 91

Micropygous, 205

Odontopleurida, 209

Parabolse, 160

Microspheric, 321

(JLcoptychius, 144

Paradoxides, 207, 211

Microspore, 332

Ogygiocaris, 208

Parasmilia. 306, 311

Migration, 33, 181

Olcostephanus, 178

Paratype, 369

Mzliolina, 325

Old Age, 27, 79, 88, 89

Parkeria, 297

Mineralization, 345, 346

Old Red Sandstone, 227

Patella, no

Miogypsina, 325

Olenellus, 207, 211

Pecopteris, 339

Modiota, 90

Olenidae, 207

Pecten, 88

Modsolopsis, 90
Mojsisovics, E., 170, 190,

Omphyma, 310
Ontogeny, n, 140, 214,

Pectinirhombs, 258
Pectunculus, 47, 86

39°

263, 287

Pelanechinus, 273

Molluscoidea, 21, 328

Oppel, A., 381, 385, 389

Peltastes, 273

Monaxonida, 318

Oppelidae, 176

Pelmatozoa, 242, 243

Monocyclic, 244, 247, 250

Ophidioceras, 139, 150

Pentacrinus, 254

Monograptus, 289, 294

Ophiocone, 139, 150, 170

Pentadactyle, 234

Monomyarian, 68

6>/w, 84

Pentamerus, 32, 34

Monotremata, 234, 238

Opisthobranchia, no, 119

Pentreinites, 259

Montlivaltia, 311

Opisthodetic, 54

Perforata, 313, 322, 326

Morphoceras, 160

Opisthogyral, 51

Perforate, 100, 265

Mortoit'ceras, 179

Opisthoparian, 197, 207

Perignathic Girdle, 264

Miihlfddtia, 40

Oral Pole, 261

Periostracum, 113

Multituberculata, 238

Orals, 246

Peripheral Area, 132

Mural Pores, 314

Orbiculoidea, 25

Periproct, 264

Murex, 116

Orbitoides, 324, 383

Perisphinctes, 147, 160 ,

Mutation, 269, 359

Orbitolina, 326, 327

177

A/>a, 91, 92

Orbitolites, 325, 382

Perissodactyla, 235, 238

^/y concha.^ 90

Orbitremites, 260

Peristome, 98, 264

Myophoria. 79

Orbulina, 323

Perna, 86

Myophoric Lamina, 56, 80

Ornithella, 3, 21, 27, 41

Peronidella, 319

412

PAL/EONTOLOGY

Peronoceras, 175

Potamides, 113

Radial Symmetry, 242

Perradii, 244

Prae-astartacea, 79

Radiative Adaptation, 231

Petalocrinus, 254

Prae-heterodonta, 77

Radiolaria, 319

Petaloid, 268

Primary Plates, 263

Radiole, 240, 261, 265

Petrifactions, 335

Primates, 238

Rafinesquina, 28

Phacops, 2ii

Prionotropidae, 179

Raphidone>na, 319

Pharetrones, 319
Philhpsastrcea, 310

Priority, Law of, 361
Proboscidea, 235, 238

Rasttites, 295
Recapitulation, n

Phillip sia, 208
Pholadomya, 70, 91

Productus, 29, 30, 31
Proetidae, 207, 212

Regularia, 270
Rejuvenation, 308

Pholas, 92, 348

Profile View, 289

Remopleurides, 207

Phosphatization, 346, 347

Progressive Characters,

Reniform Elevations, 31

Phragmoceras, 150

75, 20^

Replacement, 318

Phragrnocone, 149, 182

Prolecanites, 171

- Reptilia, 237, 237

Phragmoteuthis, 189

Pronorites, 171

Requienia, 81

Phyllocarida, 215, 217

Pro-Ostracum, 183

Resilium, 52

Phylloceras, 172, 178, 181

Proparia, 197, 210

Resorption. 12, 105, 160,

Phyllode, 277

Protaspis, 214

263 , ,

Phyllograptus, 288

Protegulum, n, 12

Retioliiidae, 294

Phylloidal, 172

Protocardia, 80

Reversion, 142

Phyllopoda, 215, 216

Protocaris, 216

Rhipidocrinus, 254

Phylogeny, n, 74, 205, 287

Protoconch, 124, 135

Rhipidoglos^a, in

Phylum, 21, 355

Protosepta,, 301

Rhodocrinus, 254

Physa, 121

Protospongia, 318

Rhombifera, 257 258

Pinacoceratidae, 172

Protozoa, 319

Rhynchocephalia, 237

Pinna, 86

Protremata, 22, 25

Rhyncl.onella, 14, 35, 37

Pinnules, 246

Provinces, Zoological, 213

Rhynchotreta, 36

Pisces, 236

Psaronius, 339

R line I la, 103, 115

Pisocrinus, 250

Pseudobelus, 187

Rostrum, 133, 153, 154,

Placocystis, 258

Pseudo-ceratites, 163, 179

359, 168, 176, 266

Placoid Scales, 222

Pseudodiadema, 273

Rotalia, 323

Placunopsis, 88

Pseudolioceras, 175

Rotiform, 165

Planktic, 145

Pseudontonotis, 64, 86

Rotula, 276

Planorbis, 107, 121
Planorbulina, 323

Psiloceras, 173
Pteria, t6

Rowe, A. W., 269 279
Rudistes, 80

Planulate, 165, 176

Pteridophyta, 332, 334

Rugosa, 302, 309

Plastron, 268, 277

Pteridosper niece, 340

Rustella, 24

Platidia, 40

Pterinopecten, 89

Platyceras, 113

Pterodactyls, 237

Platycrinus, 251

Pterophyllum, 341

Saccawmina, 326, 327

Plcsiosaurus, 233

Pteropoda, 120, 121

Saccocoma, 255

Plesiotype, 370

Pterosauria, 233

Saddles, 133, 161, 163

Pleura, 191, 193
Pleuracanthodii, 236

Pterygotus, 218
Ptilophyllum, 341

Salenia, 273
Sauropterygia, 233, 237

Pleuracanthits, 230

Ptygmatis, 115

Saxicava, 91

Plcurograptus, 291

Pugnax, 36

Seal aria, 112

Pleuronna, 91

Pulchellidae, 178

Scalariform View, 289

Pleuropterygii, 236

Pulmonata, no, 121

Sctiphandef , 119

Pleurototna, 1 1 7

Pulp Cavity, 225

S cap he I la, 117

Pleurotomnria, in

Pulvinulina, 323

Scaphites, 144, 16 1, 169,

Plicatula, 89

Pupa, 121

179

Plocoscyphia, 319

Purbeck Marls, 217

Scap! itoid, 44, 176

Podia, 261

Pygaster, 273, 275

Schellivientlla, 28, 30

Polymorphidae, 174
Polyp, 296, 303. 304

Pygidiunt, 191, 197
Pygope, 39

Schizaster, 278
Schizodont, 57

Polypary, 283

Pyrula, 116

Schizodus, 79

Polystomella, 324

Schizoneura, 338

Polyzoa, 328

Schlcenbachia, 179

Popanoceras, 171

Quenstedt, F.A., 190, 385,

Schlotheimia, 173

Porantbonites, 35

389

Sclerites, 192

Porcellanea, 322

Quenstedtoceras, 177

Scorpiones, 215, 219

Pore- Pairs, 261

Scutella, 276

Porifera, 316

Seed, 333

Porpoceras, 175

Radial Line, 128, 168

Segmental Furro , 194

Posterior Canal, 103

Radials, 244

Setninula, 43

INDEX

413

Sepia, 187

Stepheoceratidae, I76»

Thecosmilia, 311

Septa, 5, 124, 299

Stereom, 241

Thelodus, 222, 227

Septal Necks, 125
Septal Plates, 34

Stigmaria , 338
Stipe, 283

Theromorpha, 237
Thetironia, 80

Septicarina, 175

Stomechinus, 273

Thracia, 91

Septicostae, etc., 147

Strepielasma, 310

Tiarechinus, 272

Serpenticone, 143

Streptoneura, 109, ito

Tinopoms, 323

Shell-growth, 192

Stricklandian Code, 365

Tissotia, 179

Shumaraia, 209

Stric/tlandinia, 34

Tmetoceras, 175

Sicula, 282

Stringocephalus, 38

Tooth Structure, 225, 226

Side-plates, 260

Stroiaatoporoidea, 297

Topotype, 370

Sigillaria, 338, 342, 343

Strombidas, 115

Tornoce as, 153, 171

Sigtllariostrobus, 338

Stropheodonta, 28

Toucasia, 82

Silicification, 346

Strophomena, 28

Toxoceras, 142, 178

Silicispongiae, 318

Strophonella, 28

Toxocone, 142, 178

Smupalliate, 64

Stvliola, 120

Trachyceratidae, 171

Sinus, 33, 38

Stvlonurus, 218

Trapezium, 84

Siphonia, 318

Sub-petaloid, 268

Trctnatis, 25

Siphonostome, 102, 113
Sif>honotreta, 25

Suess, E., 169, 390
Sulcus, 38

Triarthrus, 200
Trigonia, 53, 79

Siphuncle, 126

Supplemental Skeleton,

Trigonocarpum, 340

Skin-teeth, 221, 222

323

Trigonosemus , 40, 41

Slit, Marginal, 96

Suture, 98, 126, 161, 163,

Trilobita, 191, 215, 397

Sniilotrochus, 311

244

Trimerellidae, 24

Smith, William, 17

Swinnerton, H. H., 206,

Tritnerus, 209

Solarium, 112

220

Trinucleus, 203, 209, 212

Solenocheilus, 150

Sycum, 117

Triplesia, 35

Solenopora, 334

Synapticulae, 313

Triserial, 273

Solnhofen Stone, 255, 297

Synaptychus, 161

Trivia, 105, 115

Somites, 191

Synonym, 361

Trivial Name, 356

Sonnininae, 175

Syntrophia, 35

Trivium, 277

Sowerby, J., and J. de C.,

Syntype, 369

Trochus, in

374

Syringopora, 314

Truncatulina, 323

Spatangoidea, 277

Syringothyris, 42

Tubicola, 329

Species, 5, 22, 355

Tube Feet, 261

Spermaphyta, 332, 340

Tubercles, 133, 147, 261

Sphcerexochus, 211

Tabulae, 302

Tuberculate, 167

Sphaerocone, 140, 150, 174

Tachygenesis, 308

Turbinate, 112

Spfueronites, 259

Tancredia, 79

Turbinolia, 311

Sphenodiscus, 179

Taxocrinus, 254

Turbinolopsis, 309

Sphenophyllales, 335, 337

Taxodont, 52

Turbo, 112

Sphenopteris, 340

Taxonomy, 73

Turreted, 99

Spicules, 316

Teleostei, 228

Turricone, 144

Spiracle, 260

Teleostomi, 236

Turrili'es, 169, 179

Spire, 08

Tellina, 84

Turritella, 97, 112

Spirifer,^, 41, 42
Spin/erina, 42, 43

Telotremata, 22, 35
Tentaculites , 120

Type Specimens, 369
7>/>/hj, 116

Spiroceras, 142

Terebratella, 40

Spiroloculina, 325

Terebralula, 2, 39, 45

Spirula, 124, 188

Terebratulina, 39

Umbilical Area, 132

Spiruli rostra, 188

Teredo, 92, 348

Umbilicus, 100, 127, 129

Spondylium, 34

Test, 261

Unguiculata, 238

Spondylus, 89 - •"
Sporadic, 181

Tethys, 172
Tetrabranchiata 147

Ungulata, 235, 238
Unicardium, 82

Sporangia, 332

Tetracidaris, 271

Unio, 77

Spores, 332,

Tetracoralla, 302, 309

Uniserial, 273

Sporophylls. 332

Tetragraptus, 286

Univalve, 95

Sporophyte, 332

Te'rameroceras, 136, 150

Spumellaria, 320

Tctrarhynchia, 37

Squamata, 237

Tetraxonida, 318

Variation, 355

Static Characters, 205

Textulau'a, 323, 326, 327

Varices, 103, 104

Staurocephalus, 211

Thallophyta, 331

Variety, 356, 359

Stegocephalia, 230, 237

Thantnastrcea, 313

Venericardia, 84

Stelleroidea, 243) 278

Theca, 244, 283, 302

Venter, 132

Stephanophyllia, 313

Thecidea, 14, 15

Ventriculites, 319

414

PALEONTOLOGY

Vtnvs, 85
Verruculina, 318
Vertebrae, 228, 233
\7ertebraria, 340
Vertebrata, 221
Virgula. 288
Visceral Cavity, 302
Vivipara, 112
Voltzia, 342
Volutilithes, 117

Walchia 342

Wea*den Shales, 217, 227
Webbtna, 323
Wenlock Beds, 32, 42, 289,
310
Whorl, 97, 129, 165
Williamsonia, 348
Wilsonia, 36
Wood, E. M. R., 290, 294,
314, 395, 396
Woodocrtnus, 254

Xipheroceras, 134, 174

Xiphosura, 215, 218
Yoldia, 76

Zatnttes, 341
Zaphrentis, 298, 310
Zellanta, 39, 40
Zoantharia, 309
Zoarium, 328
Zone, 19, 20, 269, 296,
Zooecium, 328
Zygospira, 42

BILLING AND SONS, LTD., PRINTERS, GUILDFORD, ENGLAND

14 DAY USE

RETURN TO DESK FROM WHICH BORROWED

EARTH SCIENCES LIBRARY

This book is due on the last date stamped below, or

on the date to which renewed.
Renewed books are subject to immediate recall.

General Library