Essentials of Modern Optical Fiber

Communication 2nd Edition Reinhold

Noé (Auth.)
Visit to download the full and correct content document:
https://textbookfull.com/product/essentials-of-modern-optical-fiber-communication-2n
d-edition-reinhold-noe-auth/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Elements of Optical Networking: Basics and Practice of

Glass Fiber Optical Data Communication 2nd Edition
Brückner

https://textbookfull.com/product/elements-of-optical-networking-
basics-and-practice-of-glass-fiber-optical-data-
communication-2nd-edition-bruckner/

Essentials of business communication Guffey

https://textbookfull.com/product/essentials-of-business-
communication-guffey/

Free Space Optical Communication Jain

https://textbookfull.com/product/free-space-optical-
communication-jain/

Polymer Optical Fiber Bragg Gratings-Fabrication and

Sensing Applications 1st Edition Ricardo Oliveira

https://textbookfull.com/product/polymer-optical-fiber-bragg-
gratings-fabrication-and-sensing-applications-1st-edition-
ricardo-oliveira/
Essentials of Business Communication Mary Ellen Guffey

https://textbookfull.com/product/essentials-of-business-
communication-mary-ellen-guffey/

Free Space Optical Communication 1st Edition Hemani

Kaushal

https://textbookfull.com/product/free-space-optical-
communication-1st-edition-hemani-kaushal/

Plastic Optical Fiber Sensors: Science, Technology and

Applications 1st Edition Marcelo M. Werneck (Editor)

https://textbookfull.com/product/plastic-optical-fiber-sensors-
science-technology-and-applications-1st-edition-marcelo-m-
werneck-editor/

Essentials of Human Communication, 11th Edition Joseph

A. Devito

https://textbookfull.com/product/essentials-of-human-
communication-11th-edition-joseph-a-devito/

Essentials of Human Communication 11th Edition Joseph

A. Devito

https://textbookfull.com/product/essentials-of-human-
communication-11th-edition-joseph-a-devito-2/
 Reinhold Noé

Essentials
of Modern
Optical Fiber
Communication
Second Edition

123
Essentials of Modern Optical Fiber Communication
Reinhold Noé

Essentials of Modern Optical

Fiber Communication
Second Edition

123
Reinhold Noé
Faculty of Computer Science, Electrical
Engineering and Mathematics, Institute
for Electrical Engineering and Information
Technology
Paderborn University
Paderborn
Germany

ISBN 978-3-662-49621-3 ISBN 978-3-662-49623-7 (eBook)

DOI 10.1007/978-3-662-49623-7

Library of Congress Control Number: 2016935217

© Springer-Verlag Berlin Heidelberg 2010, 2016

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer-Verlag GmbH Berlin Heidelberg
Preface

This book covers important aspects of modern optical communication. It is inten-

ded to serve both students and professionals. Consequently, a solid coverage of the
necessary fundamentals is combined with an in-depth discussion of recent relevant
research results.
The book has grown from lecture notes over the years, starting 1992. It
accompanies my present lectures on Optical Communication A (Fundamentals),
B (Mode Coupling), C (Modulation Formats) and D (Selected Topics) at Paderborn
University in Germany.
I gratefully acknowledge contributions to this book from Dr. Timo Pfau,
Dr. David Sandel and Prof. Dr. Sebastian Hoffmann.

v
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 Optical Waves in Fibers and Components. . . . . . . . . . . . . . . . . . . . 3
2.1 Electromagnetic Fundamentals. . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1.1 Maxwell’s Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1.2 Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.1.3 Wave Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1.4 Homogeneous Plane Wave in Isotropic
Homogeneous Medium . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1.5 Power and Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2 Dielectric Waveguides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.2.1 Dielectric Slab Waveguide . . . . . . . . . . . . . . . . . . . . . . . 29
2.2.2 Cylindrical Dielectric Waveguide . . . . . . . . . . . . . . . . . . 38
2.3 Polarization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.3.1 Representing States-of-Polarization . . . . . . . . . . . . . . . . . 52
2.3.2 Anisotropy, Index Ellipsoid . . . . . . . . . . . . . . . . . . . . . . 58
2.3.3 Jones Matrices, Müller Matrices . . . . . . . . . . . . . . . . . . . 65
2.3.4 Monochromatic Polarization Transmission . . . . . . . . . . . . 82
2.3.5 Polarization Mode Dispersion . . . . . . . . . . . . . . . . . . . . . 92
2.3.6 Polarization-Dependent Loss . . . . . . . . . . . . . . . . . . . . . . 100
2.4 Linear Electrooptic Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
2.4.1 Phase Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
2.4.2 Soleil-Babinet Compensator . . . . . . . . . . . . . . . . . . . . . . 110
2.5 Mode Coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
2.5.1 Mode Orthogonality. . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
2.5.2 Mode Coupling Theory . . . . . . . . . . . . . . . . . . . . . . . . . 119
2.5.3 Codirectional Coupling in Anisotropic Waveguide. . . . . . . 122
2.5.4 Codirectional Coupling of Two Waveguides . . . . . . . . . . . 129
2.5.5 Periodic Codirectional Coupling . . . . . . . . . . . . . . . . . . . 135
2.5.6 Periodic Counterdirectional Coupling . . . . . . . . . . . . . . . . 140

vii
viii Contents

2.6 Differential Group Delay Profiles . . . . . . . . . . . . . . . . . . . . . . . . 142

2.6.1 DGD Profiles and Discrete Mode Coupling . . . . . . . . . . . 142
2.6.2 Polarization Mode Dispersion Compensation. . . . . . . . . . . 148
2.6.3 Chromatic Dispersion Compensation . . . . . . . . . . . . . . . . 154
2.6.4 Fourier Expansion of Mode Coupling . . . . . . . . . . . . . . . 160
2.6.5 DGD and PDL Profiles Determined
by Inverse Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
2.7 Nonlinearities in Optical Fibers . . . . . . . . . . . . . . . . . . . . . . . . . 169
2.7.1 Self Phase Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . 170
2.7.2 Cross Phase Modulation . . . . . . . . . . . . . . . . . . . . . . . . . 181
2.7.3 Four-Wave Mixing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
3 Optical Fiber Communication Systems . . . . . . . . . . . . . . . . . . . . . . 189
3.1 Standard Systems with Direct Optical Detection . . . . . . . . . . . . . 189
3.1.1 Signal Generation, Transmission, and Detection . . . . . . . . 189
3.1.2 Regeneration of Binary Signals . . . . . . . . . . . . . . . . . . . . 206
3.1.3 Circuits and Clock Recovery. . . . . . . . . . . . . . . . . . . . . . 214
3.2 Advanced Systems with Direct Detection . . . . . . . . . . . . . . . . . . 222
3.2.1 Photon Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
3.2.2 Noise Figure of Optical Amplifier . . . . . . . . . . . . . . . . . . 228
3.2.3 Intensity Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . 232
3.2.4 Receivers for Amplitude Shift Keying . . . . . . . . . . . . . . . 236
3.2.5 Receivers for Differential Phase Shift Keying . . . . . . . . . . 241
3.2.6 Polarization Division Multiplex . . . . . . . . . . . . . . . . . . . . 256
3.3 Coherent Optical Transmission . . . . . . . . . . . . . . . . . . . . . . . . . 262
3.3.1 Receivers with Synchronous Demodulation. . . . . . . . . . . . 262
3.3.2 Carrier Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
3.3.3 Receivers with Asynchronous Demodulation. . . . . . . . . . . 283
3.3.4 Laser Linewidth Requirements . . . . . . . . . . . . . . . . . . . . 287
3.3.5 Digital Coherent QPSK Receiver. . . . . . . . . . . . . . . . . . . 294
3.3.6 Digital Coherent QAM Receiver . . . . . . . . . . . . . . . . . . . 306
3.3.7 Other Modulation Schemes. . . . . . . . . . . . . . . . . . . . . . . 324

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
Chapter 1
Introduction

At the end of the 1970s, telecom carriers started to lay optical fiber between telecom
exchange offices, and coaxial cable for electrical data communication was no longer
deployed. The performance of optical fiber communication has since then grown
exponentially, very much like Moore’s law for the complexity of electronic circuits.
In the electronic domain, rising clock speeds, miniaturization of feature sizes, and
chip size increase along two, maybe soon along the third dimension, are con-
tributing to this truly impressive growth. The performance of optical communica-
tion is determined by clock speed as offered by a state-of-the-art electronic
technology, availability of several or if needed many fibers in one cable, multiple
optical channels carried on a single optical fiber by means of wavelength division
multiplex, and recently the transmission of several bits per symbol.
The economic and societal impact is dramatic: Optical fiber communication is a
key enabler of the worldwide web, of e-mail and of all but local telephone con-
nections. The technically exploitable fiber bandwidth is roughly 10 THz, orders of
magnitude higher than in other media. Fiber attenuation is extremely small: After
100 km of fiber there is still about 1 % of the input power left. Optical amplifiers,
with 4 THz bandwidth or more, overcome fiber loss so that transoceanic trans-
mission is possible without intermediate signal regeneration.
Around the year 2000, in the so-called dotcom era, growth rates of information
exchange of about one order of magnitude per year were forecast. This triggered
massive investments and resulted in the founding of many new companies in a short
time. A significant part of that investment was lost, while achieved technical pro-
gress remains available at large. The telecom industry has consolidated since then
because investments make sense only if customers pay them back. Of course,
customers don’t want to spend a significant part of their household budget for
communication, even though available bandwidth has grown by more than two
orders of magnitude thanks to DSL technology.
But today’s communication does indeed grow by a factor of 1.4 per year or so.
Private communication such as music downloading, video portals, personal web-
sites and of course also the ever more complex and video-laden media and
© Springer-Verlag Berlin Heidelberg 2016 1
R. Noé, Essentials of Modern Optical Fiber Communication,
DOI 10.1007/978-3-662-49623-7_1
2 1 Introduction

enterprise websites are responsible for this, along with video telephony services,
drastically increasing usage of the internet in developing countries, and so on. As a
consequence there is healthy business. In contrast, revenues increase only on a
single-digit percent scale annually. The quantitative growth is entertained by the
technical and productivity progress.
With the rather conservative spending pattern of end users in mind, telecom
carriers want to preserve their enormous investments in fiber infrastructure, and to
use newly deployed fiber most economically. Multilevel modulation schemes,
including the use of two orthogonal polarization modes, are needed to exploit fibers
optimally. At the same time, phase modulation increases noise tolerance. Recent
research and development places special emphasis on these issues, and so does this
book.
Understanding fibers requires a knowledge of dielectric waveguides and their
modes, including polarizations. Chapter 2 is therefore devoted to wave propagation
in ideal and nonideal optical waveguides, also exhibiting polarization mode dis-
persion and polarization-dependent loss, to mode coupling, electrooptic compo-
nents and nonlinear effects in silica fibers. Most optical components and
transmission effects are based on these features.
Chapter 3 discusses optical transmission systems of all kinds. The simplest are
standard intensity-modulated direct-detection systems. Their reach can be dramat-
ically extended by optical amplifiers, the theory of which is thoroughly described.
Performance is enhanced by binary and quadrature phase shift keying with inter-
ferometric detection. Symbol rate can be doubled by polarization division multi-
plex. The same is possible also with coherent optical systems. But these can as well
detect signal synchronously, which again increases performance. The principle is
that the received signal and the unmodulated signal of a local laser are superim-
posed. The power fluctuations resulting from this interference are detected. Several
signal superpositions and detectors allow obtaining an electrical replica of the
optical field vector. Coherent optical transmission systems can therefore electron-
ically compensate all linear distortions suffered during transmission. Signal pro-
cessing and control algorithms for high-performance digital synchronous coherent
optical receivers conclude the book. Coherent transmission, which increases the
traditional fiber capacity 10- or 20-fold, has become a megatrend in optical com-
munication since 2007.
Fiber-to-the-home services can increase customer data rates by several more
orders of magnitude and make it likely that the pressure for increased capacity at
moderate cost in metropolitan area and long haul communication will continue.
Chapter 2
Optical Waves in Fibers and Components

2.1 Electromagnetic Fundamentals

2.1.1 Maxwell’s Equations

Electromagnetic radiations obeys Maxwell’s equations

@D
curl H ¼ þJ Ampere0 s law; ð2:1Þ
@t
@B
curl E ¼  Maxwell-Faraday equation; ð2:2Þ
@t

div D ¼ q Gau
0 s law; ð2:3Þ

div B ¼ 0 Gau
0 s law for magnetism; ð2:4Þ

We take the divergence of (2.1) and obtain with div curl A ¼ 0 the

@q
div J ¼  continuity equation; ð2:5Þ
@t

It says that the current drained from the surface of a differential volume element
equals the reduction of charge per time interval (preservation of charge). The
equations can be brought into integral form, using Gauß’s and Stokes’s integral
theorems,
I ZZ  
@D @We
H  ds ¼ þ J  da ¼ þI ð2:6Þ
@t @t

© Springer-Verlag Berlin Heidelberg 2016 3

R. Noé, Essentials of Modern Optical Fiber Communication,
DOI 10.1007/978-3-662-49623-7_2
4 2 Optical Waves in Fibers and Components
RR RR
(I ¼ J  da: enclosed current; We ¼ D  da: electric flux),
I Z Z 
@ @Wm
Uind ¼ E  ds ¼  B  da ¼ ð2:7Þ
@t @t
RR
(Uind : induced voltage; Wm ¼ B  da: magnetic flux),
ZZ ZZZ ZZZ
 D  da ¼ q dV ¼ Q ðQ ¼ q dV : enclosed chargeÞ; ð2:8Þ

ZZ
 B  da ¼ 0; ð2:9Þ

ZZ ZZZ 
@ @Q
I ¼  J  da ¼  q dV ¼  : ð2:10Þ
@t @t

The relations between fields and flux densities are given by the material
equations
D ¼ e0 E þ P; ð2:11Þ

B ¼ l0 H þ M: ð2:12Þ

In isotropic media electric (P) and magnetic (M) dipole moment have the same
direction as the corresponding field. Therefore the material equations simplify into

D ¼ eE ¼ e0 er E ¼ e0 ð1 þ vÞE ðv : susceptibilityÞ; ð2:13Þ

B ¼ lH ¼ l0 lr H: ð2:14Þ

But the same equations can also be applied for anisotropic media if ε (and χ) and
μ are not defined as scalars but as rank-2 tensors (matrices),

D ¼ eE ¼ e0 er E ¼ e0 ð1 þ vÞE B ¼ lH: ð2:15Þ

The material tensors are quadratic 3 × 3 matrices. In non-magnetic media, as

employed in optics, it holds lr ¼ 1, l ¼ l0 . All the same we will occasionally set l
instead of l0 in order to emphasize the analogy of treatment of magnetic and
electric field or to show that equations can be used also outside the optical domain.
The relative dielectricity constant serves to define the refractive index n through
er ¼ n2 . In vacuum it holds lr ¼ 1, er ¼ n2 ¼ 1.
Ohm’s law

J ¼ rE; ð2:16Þ

which is another material equation, relates current density and electric field.
2.1 Electromagnetic Fundamentals 5

Time-dependent signals can be expressed by summation of Fourier components

with different frequencies in the frequency domain. Therefore a complex separation
ansatz of space and time dependence such as Hðr; tÞ ¼ HðrÞejxt is particularly apt
to solve Maxwell’s equations. The physical, scalar or vectorial amplitude quantity
is simply the real part of the corresponding complex quantity. If one replaces @=@t
by jx then (2.1) becomes

curl H ¼ jxD þ J ¼ jxeE þ J with D ¼ eE; ð2:17aÞ

where we have assumed time-invariance of ε. Losses are taken into account in the
current density J, which facilitates the interpretation of Poynting’s vector. But in
optics it is often more convenient to take losses into account in a complex
dielectricity constant

e ¼ e0 er ¼ e0 ðer  jeri Þ ¼ e0 er  jr=x ¼ e  jr=x; ð2:18Þ

here defined for isotropic media. This results in a re-defined complex flux density

r 1
D ¼ eE ¼ eE  j E ¼ eE  j J; ð2:19Þ
x x

1 1 @q
div D ¼ div eE þ div J ¼ q  ¼ q  q ¼ 0: ð2:20aÞ
jx jx dt

Here (2.5) has been inserted. In (2.17a, b) the term jxD þ J is replaced by the
re-defined (by 2.19) jxD. One obtains

@D @E
curl H ¼ ¼e ¼ jxD ¼ jxeE with D ¼ eE: ð2:21aÞ
@t @t

Note that the effects of current density are duly taken into account, like in (2.17a, b).
If there are pure ohmic losses then σ is frequency-independent. Generally it
depends on frequency. Losses are characterized by r [ 0; eri [ 0. In lasers and
optical amplifiers one utilizes media which amplify electromagnetic radiation in the
optical domain, where r\0; eri \0 is valid.
The two definitions of D are based on two different usages in the literature.
While (2.21a, b) is formally (2.17a, b) in contradiction with (2.1) the current density
is correctly taken into account by the complex dielectricity constant e.
Analogously, magnetic losses can be expressed by a complex permeability
constant

l ¼ l0 lr ¼ l0 ðlr  jlri Þ: ð2:22Þ

For anisotropic media one uses complex material tensors e, l.

6 2 Optical Waves in Fibers and Components

We obtain Maxwell’s equation in complex notation

curl H ¼ jxD ¼ jxeE; ð2:21bÞ

curl E ¼ jxB ¼ jxlH; ð2:23Þ

div D ¼ div ðeEÞ ¼ 0; ð2:20bÞ

 
div B ¼ div lH ¼ 0: ð2:24Þ

With real dielectricity constant when using the other definition of the electric
flux density (dielectric displacement) it holds instead

curl H ¼ jxD þ J ¼ jxeE þ J; ð2:17bÞ

div D ¼ div ðeEÞ ¼ q; div J ¼ div ðrEÞ ¼ jxq: ð2:25Þ

2.1.2 Boundary Conditions

Normally the medium of wave propagation is not homogeneous and infinite in

pffiffiffiffi
space. For example, between air (refractive index n ¼ er  1) and silica
(n ≈ 1.46) there is a refractive index difference which must be taken into account in
the calculations. The most effective way to do this is to solve the wave propagation
equations at both sides of the boundary and to equate the two solutions with free
parameters, using the boundary conditions.
We first determine the normal boundary conditions for electric and magnetic
flux densities perpendicular to the boundary. In Fig. 2.1a the boundary region
between two media with different material properties is sketched. Let the two media
1, 2 be homogeneous and isotropic. Bottom and lid of a shallow cylinder, both with
area F, lie in media 1 and 2 with material constants l1 ; e1 and l2 ; e2 , respectively.
The unit vector n is perpendicular to the boundary plane. Let the cylinder height

Fig. 2.1 Derivation of (a) (b)

normal (a) and tangential 1 1
1 1 2
(b) boundary conditions

n
l 1
F n

2 2
h 2 2 h
2.1 Electromagnetic Fundamentals 7

h approach zero, so that its surface can be neglected. Gauß’s law for magnetism in
integral form (2.9) yields
ZZ
0 ¼  B  da ¼ ðB2  B1 Þ  n F ) B2n  B1n ¼ 0: ð2:26Þ

The normal components B1n , B2n of the magnetic flux density in direction of the
normal vector n are identical on both sides of the boundary. In other words, it must
be continuous while passing the boundary. Gauß’s law in integral form (2.8) yields
the enclosed charge. Assuming an area charge density qA , which in the boundary
itself corresponds to an infinite space charge density, the enclosed charge equals
Q ¼ qA F. In optics it usually holds qA ¼ 0. In summary it holds for the normal
components D1n , D2n of the electric flux density
ZZZ ZZ
Q¼ qdV ¼  D  da ¼ ðD2  D1 Þ  n F ) D2n  D1n ¼ qA : ð2:27Þ

For deduction of the tangential boundary conditions we look at Fig. 2.1b.

A rectangular area element have length l and height h, which tends again toward
zero. The closed-loop integral of the magnetic field is
I
H  ds ¼ l ðH2  H1 Þ  s1 ; ð2:28Þ

where s1 is the unit vector in the tangential plane parallel to a side of the rectangle.
For finite temporal changes of electric flux and current densities the right-hand side
of Ampere’s law in integral form (2.6), applied to the area element, equals zero,
since height h approaches zero. But if the boundary conductivity is infinite then
there can be an area current density JA with

Zh=2
lim Jdn ¼ JA : ð2:29Þ
h!0
h=2

Ampere’s law then yields

ðH2  H1 Þ  s1 ¼ JA  s2 ; ð2:30Þ

where s2 is the unit vector in the tangential plane that is perpendicular to s1 . If one
replaces s1 by s2  n one obtains on the left-hand side a spade product of three
vectors, which may be cyclically exchanged according to u  ðv  wÞ ¼
v  ðw  uÞ,
8 2 Optical Waves in Fibers and Components

s2  ½n  ðH2  H1 Þ ¼ s2  JA : ð2:31Þ

Since the direction of s2 in the tangential plane can be chosen at will, and JA and
n  ðH2  H1 Þ lie in the tangential plane, we may write

n  ð H 2  H 1 Þ ¼ JA : ð2:32Þ

Infinite conductivity excluded the tangential components (index t) of the mag-

netic field are continuous when passing the boundary,

H2t ¼ H1t for JA ¼ 0: ð2:33Þ

The Maxwell-Faraday equation in integral form (2.7) allows deducing in analog

fashion the continuity of the tangential electric field components in the boundary,

E2t ¼ E1t : ð2:34Þ

In (2.26) and (2.27) we have deduced the conditions for the normal components
of the flux densities. The corresponding fields are found using the material equa-
tions. Similarly, the tangential components of the flux densities can be found from
the material equations once (2.32)–(2.34) have specified the tangential field
component.
Tangential and normal boundary conditions are interrelated. To show this one
bends the area element of Fig. 2.1b to a complete cylinder wall of Fig. 2.1a. This
way the continuity of the tangential electric (magnetic) field becomes equivalent to
the continuity of the normal magnetic (electric) flux density. It is therefore sufficient
to fulfill either
• the tangential or
• the normal boundary conditions or
• the normal boundary condition for the electric flux density and the tangential
one for the electric field or
• the normal boundary condition for the magnetic flux density and the tangential
one for the magnetic field.
The other boundary conditions are then automatically fulfilled.
The homogeneous region may be limited to the immediate surroundings of the
boundary.

2.1.3 Wave Equation

We use complex notation and take losses into account in the imaginary parts of
complex material parameters e, l. The medium be isotropic so that e, l are scalars.
2.1 Electromagnetic Fundamentals 9

We take the curl operator on both sides of Maxwell-Faraday Equation (2.23) and
apply on the right-hand side the general relation curl ðFaÞ ¼ Fcurl a  a  grad F,
   
curl ðcurl EÞ ¼ jxcurl lH ¼ jx lcurl H þ H  grad l : ð2:35Þ

Ampere’s law (2.21a, b) is inserted into first term, while (2.23) is again inserted
into the second term on the right-hand side,

1
curl ðcurl EÞ ¼ x2 leE  curl E  grad l: ð2:36Þ
l

The second term on the right-hand side is roughly zero if l changes only little
within one optical wavelength. This is quite common. In optics it even holds l ¼ l0
so that grad l ¼ 0 holds. As a result we obtain

curl ðcurl EÞ ¼ x2 leE: ð2:37Þ

According to (2.20a, b) and with div ðFAÞ ¼ Fdiv A þ A  grad F we can write

0 ¼ div ðDÞ ¼ ediv E þ E  grad e: ð2:38Þ

We insert into (2.37) the Laplace operator DA ¼ grad ðdiv AÞ  curl ðcurl AÞ
and (2.38) solved for div E,
 
1
DE  grad ðdiv EÞ ¼ DE þ grad E  grad e ¼ x2 leE: ð2:39Þ
e

In vacuum and other homogeneous media (i.e., e, l are position-independent),

but with sufficient accuracy also in slightly inhomogeneous media it holds
ðgrad eÞ=e ¼ 0. This results in a simplified wave equation for the electric field,

DE þ x2 leE ¼ 0: ð2:40Þ

Due to the symmetry of Maxwell’s equations one can derive in analog fashion
for the magnetic field

DH þ x2 leH ¼ 0: ð2:41Þ

The vectorial wave Eq. (2.40) tells only the relation between space and
time-dependence of the wave amplitude. However, the direction of the field vector
is yet unclear. Once (2.40) is solved one may choose a tentative arbitrary vector
direction E. Then one calculates H through the Maxwell-Faraday equation. Finally
H is inserted into Ampere’s law and one obtains a usually modified E which is the
correct solution. Instead of this complicated procedure one may start with certain
10 2 Optical Waves in Fibers and Components

assumptions (Sect. 2.1.4) or may eliminate a degree of freedom of the field vector
(Sect. 2.3.2). The same holds for solutions of (2.41). An elegant possibility for
“direct” solution of Maxwell’s equations are electromagnetic potentials.
We assume now a nonmagnetic medium (l ¼ l0 ), insert the

1
c ¼ pffiffiffiffiffiffiffiffiffi speed of light in vacuum ð2:42Þ
e0 l0

and the definition of the refractive index

pffiffiffiffi
n¼ er ; n ¼ ðer Þ1=2 ð2:43Þ

and write as an example (2.40) as

n2 @ 2 E n2
DE ¼ ¼ x2 2 E: ð2:44Þ
c @t
2 2 c

The wave equation is often solved numerically. Knowledge of the

position-dependence of e or n and, if applicable, of an incident field E, is required.

2.1.4 Homogeneous Plane Wave in Isotropic

Homogeneous Medium

We investigate wave propagation in an isotropic, homogeneous medium and write

Ampere’s law (2.21a, b) and the Maxwell-Faraday equation (2.23) in cartesian
coordinates

@H z @H y @E
 ¼e x; ð2:45Þ
@y @z @t

@H x @H z @E y
 ¼e ; ð2:46Þ
@z @x @t
@H y @H x @E
 ¼e z; ð2:47Þ
@x @y @t

@E z @Ey @H x
 ¼ l ; ð2:48Þ
@y @z @t

@E x @E z @H y
 ¼ l ; ð2:49Þ
@z @x @t
22 2 Optical Waves in Fibers and Components

Fig. 2.3 Reflection and transmission as a function of incidence angle a1 from air to glass (top)
and from glass to air (bottom). E parallel (-) or perpendicular (- -) to incidence plane. Brewster
angle (○) and critical angle for total reflection (□) are marked. At the Brewster angle the phase
(=the argument) of q11 jumps by π (bottom right). Air: n = 1; glass: n = 1.46

Fig. 2.4 Reflection and transmission of homogeneous plane wave at multiple dielectric layers

incidence plane. Note that most of the multiple reflections at and between the layers
in
are not depicted in the drawings. Use the same reference point A for the waves E1a ,
out in
E2b , E3a independent of tilt angle a1 .
Calculate H and R for m ¼ 2.
Another random document with
no related content on Scribd:
The muscle-scars in this family (Fig. 323, A, B) are most remarkable for the
development of the so-called “crescent,” (q.r.s.) which skirts the posterior
margin of both valves as a sub-cardinal impression. It is believed to be the
trace of a strong post-parietal muscular wall, analogous in position to that of
Lingula. The three pairs of “lateral” muscle-scars in the latter genus seem to
be represented by the “terminal” (s) and “lateral” (r) scars on the crescent of
the Trimerellidae. A pair of “transverse” scars (t) occurs in each valve between
the “terminals” and the antero-lateral edge of the “platform” (j). “Cardinal” (v),
“sub-cardinal” (w), and “umbo-lateral” (x) scars also occur. The median
impression which covers the “platform” (j) consists of a central, lateral, and
usually an anterior pair of scars; and the impressions of the genital organs,
according to Davidson and King, lie medianly posterior to the “platform.” The
“platform” itself is a more or less conspicuous central calcareous elevated area
occurring in each valve, but most developed in the dorsal; in some cases it is
double-chambered with tubular cavities (“platform vaults,” Fig. 323, A, B, k), in
others it is more or less solid. It appears to have originated through a posterior
shifting of the central muscular bands, that they might be inserted behind the
liver; at the same time a deposition of shelly material, to form fulcra to work the
heavy valves, took place at these points. The tunnelling-out of the platform
was probably due to the continual pressure of the lobes of the liver. The
division of the umbonal cavity into definite chambers in Monomerella, and to a
less extent in other members of this family, appears, according to Davidson
and King, to have been caused by pressure of the ovarian lobes.
In connexion with the foregoing remarks on the development of the
“platform,” it may be mentioned that the paths along which the muscle-bands
move, as the shell of Brachiopods increases in size, are marked by elongated
scars, and often by shelly deposits; and when the members of a muscle-pair
come into juxtaposition these shelly deposits (which act as fulcra for the
muscles) combine, and by the growth of the shell form a septum, as in the
case of the median septum of Lingulepis.
The Obolidae show some important features in the internal impressions.
Obolella crassa (Hall) may be taken as a well-known type of the family. In this
species a pair of small scars, one on each side of the pedicle-groove, lies
close under the hinge line in the ventral valve. There is also a well-marked
scar for the insertion of the pedicle-muscle at the end of the pedicle-groove. A
pair of much elongated lateral impressions extending forward from the
“cardinals” may be homologous with the “laterals” of Lingula; and the two small
central scars between them may be compared with the “centrals” of Lingula
which are in a somewhat similar position. In the dorsal valve of O. crassa a
pair of “cardinals” is found, and on each side of a low median rounded ridge
are two small “central” scars. Indistinct “lateral” scars arise close to or in the
central area, and diverge anteriorly.
Sometimes a great concentration of muscle-scars occurs round the foramen
in the ventral valve, as in Siphonotreta.
As regards the minute structure and composition of the shell in the
Ecardines, we find that the Lingulidae and Discinidae have their shell
composed of alternating layers of phosphate of lime and a corneous
substance; the former layers are pierced by microscopic canals. The Craniidae
have calcareous shells traversed by tubules, which divide into many fine
branches near the external surface; a thin periostracum covers the exterior.
The Trimerellidae have heavy thick calcareous shells, for which they required
the previously-described elaborate arrangement of muscles to open and shut
them.

II. TESTICARDINES

External Characters

It is to this division that the great majority of the Brachiopoda belong; and
the diversity of form, of ornamentation, and of internal characters is
correspondingly greater than in the Ecardines.
A transversely or longitudinally oval shape of shell is the commonest; but
sometimes it is triangular, as in Rhynchonella (Fig. 327), or bilobed, as in
Pygope (= Terebratula diphya). The ventral valve is usually more convex than
the dorsal, and the former may be prolonged into a tube by the accelerated
growth and infolding of the anterior and lateral margins, producing a very
abnormal form (Proboscidella). The external surface of the valves is frequently
ornamented with more or less prominent radiating ribs; and fine concentric
growth-lines are commonly shown, and may be developed into coarse ridges
or wrinkles, particularly in old individuals. The members of the family
Productidae are usually furnished with tubular spines, which are sometimes of
great length, and served to anchor the free shells in the mud, or were twisted
round Crinoid stems and similar objects.
In the ventral valve of many genera there is a median sinus, with a
corresponding fold in the dorsal valve, and rarely vice versâ; sometimes the
fold and sinus are double.
The hinge line is either curved or straight, and the valves are articulated by
means of a pair of “hinge-teeth” (Fig. 329, t) in the ventral valve, which fit into
corresponding sockets in the opposite valve. Some genera have the teeth very
rudimentary, or have lost them altogether. The teeth are frequently supported
by “dental plates,” and the sockets by “socket plates” (e.g. Conchidium, Figs.
324, 325). A few genera with a long hinge line have the whole of it
denticulated (Stropheodonta). In the dorsal valve medianly close under the
hinge line is a shelly protuberance—the “cardinal process”—to which the
diductor muscles are attached. It is sometimes of great length and forked
(Stringocephalus, Fig. 326), or tripartite, or even quadripartite; but in
Rhynchonella and some other genera it is rudimentary.

Fig. 324.—Conchidium
galeatum. Wenlock
Limestone.

Fig. 325.—Conchidium galeatum.

Transverse section. d, Dorsal
valve; d.s, dorsal septum; s,
socket plate; v, ventral valve; v.s,
ventral septum; d.p, dental plate.
 Fig. 326.—Stringocephalus Burtini. (Modified from Woodward.)
Devonian. A, Interior of dorsal valve. B, Side view of interior
of shell; a, adductor (= occlusor) scars; c, crura; c.p, cardinal
process; d.s, dorsal septum; h.p, hinge plate; l, brachial loop;
s.p, shelly processes; t.s, dental sockets; v.s, ventral
septum.
A “hinge area” (Fig. 334, c.a) is often present on one or both valves, and
may be of great size, as in Clitambonites, but in Productus it is wholly absent.
In those genera that possess it a triangular fissure—the “deltidial fissure”—
frequently traverses it on both valves; in the dorsal valve the fissure is merely
the space between the dental sockets, and may be occupied by the cardinal-
process (Fig. 334, C) or covered by a shelly plate—the “chilidium.” In the
ventral valve it gives passage to the pedicle, and may be partly or entirely
closed by a similar plate (Fig. 334, d) known as the “pseudo-deltidium,”
especially large in Clitambonites, or remain open (Orthis). This pseudo-
deltidium is a primitive character, and arises in an early stage of the
development as a shell-growth on the dorsal side of the animal, becoming
attached to the ventral valve subsequently. The pedicle in many genera
passes out through a special foramen in the beak of the ventral valve; and its
proximal portion is often embraced by a pair of small plates—the deltidial
plates or “deltidium”—which are formed on lateral extensions of the ventral
mantle lobe, according to Beecher. These plates lie on each side of the
pedicle, or grow round and unite in front of it (Rhynchonella, Fig. 327), or
constitute merely its anterior border (Terebratula, Fig. 328). In some cases this
foramen becomes closed in old age.
 Fig. 327.—Rhynchonella
Boueti. (Cornbrash.) d,
Deltidium; f, foramen.

Fig. 328.—Terebratula
sella. (Lower
Greensand.) d,
Deltidium; f, foramen.
The dorsal valve in a few cases has its beak perforated by a foramen—the
“visceral foramen.” This foramen is in no way connected with the pedicle
foramen, but points perhaps to the existence in the early Testicardinate genera
of an anal aperture. In Athyris concentrica (Devonian) this foramen is
connected internally with a cylindrical tube, which extends longitudinally to
about one-third the length of the valve. In Centronella the aperture in the
cardinal plate is rounded and complete; and in Strophomena and its allies the
opening lies between the cardinal processes. If this feature is correctly
interpreted, it suggests a retrogression of the group since Palaeozoic times not
only in numbers, but in structure; and other evidence points the same way.

Internal Characters
The interior of the shell is sometimes more or less divided up by septa. A
median septum occurs in one or both valves of many genera as a low ridge or
strongly developed partition (Waldheimia, Fig. 329, ss; and Stringocephalus,
Fig. 326, B, v.s). Conchidium (Fig. 325) has its dental plates of great size, and
uniting to form a V-shaped chamber or “spondylium,” supported by a median
double septum; and by means of these with a pair of septa and the large
socket-plates in the dorsal valve the interior of the shell of this genus is divided
up into several chambers.
The interiors of several other genera are somewhat similarly divided up.

Fig. 329.—Waldheimia (Magellania) flavescens. A, Interior of

ventral valve: a, adductor scars; v.a, ventral adjustors; d,
divaricators; a.d, accessory divaricators; p, peduncular
muscle; dm, deltidium; f, foramen; t, teeth. B, Interior of
dorsal valve: a.a, anterior adductor (occlusor) scars; a.p,
posterior adductor (occlusor) scars; c.p, cardinal process;
cr, crura; d.s, dental sockets; hp, hinge-plate; l, brachial
loop; ss, septum. (After Davidson.)
In the Carboniferous genus Syringothyris two special plates, situated
between the dental plates, are rolled into an incomplete tube, so as to enclose
probably the anal extremity of the alimentary canal; and in several genera a
sub-umbonal “cardinal plate” is present, which is perforated (Athyris) or slit in
some cases for the passage of the anal tube.
For the support of the fleshy “spiral arms” the calcareous structures forming
the “brachial apparatus” are of two main types—(1) the loop type; (2) the
spiral-cone type. In the Strophomenidae no special calcareous support seems
to have been usually present (Fig. 334), though in some species of Leptaena
spirally-grooved elevated areas supported the fleshy arms; in the Productidae
it is probable that the ridges enclosing the “reniform impressions” (Fig. 333, i)
served for a similar purpose.
 The Terebratulidae show the “loop type” of brachial apparatus. In
Waldheimia (Fig. 329), which may be taken as an example, we notice first in
the dorsal valve the “crura” (cr), from which arise the two “descending
branches” which run forwards and then are bent back to form the “ascending
branches” which are united by the “transverse band.” In some genera the
“ascending branches” may be reduced to mere points, and the “transverse
band” become a median vertical plate; the “crura,” too, may be fused so as to
form a “crural band”; and the “descending branches” may be connected by a
cross band—the “jugal band.” In Stringocephalus (Fig. 326, l, s.p) the loop is
furnished on its inner edge with radiating processes; and in Argiope the loop is
simple, not reflected, and fused with marginal septa; while in the Thecidiidae it
is more or less fused with the shell itself, and with the mass of calcareous
spicules secreted by the mantle.
The “spiral-cone type” of brachial apparatus is found in the Spiriferidae,
Atrypidae, and Koninckinidae, and consists of two spirally-enrolled calcified
lamellae, forming two cones with their apices directed laterally (Spirifera, Fig.
330), or towards the interior of the dorsal valve (Atrypa, Fig. 332), or towards
each other (Glassia); or forming two flat spirals in the same plane
(Koninckinidae). A “jugal band” is generally present, but varies much in
position, and in some genera has complicated posterior processes.
The Rhynchonellidae have no loop or spiral cones, but merely a pair of
short “crura.”

Fig. 330.—Spirifera striata. (Carboniferous

Limestone.) Showing brachial spires.
The principal modifications in the attachments of the muscles in the
Testicardines are illustrated by Productus giganteus (Fig. 333), Leptaena
rhomboidalis (Fig. 334), and Waldheimia flavescens (Fig. 329).
In Productus (Fig. 333) we see in the ventral valve a pair of dendritic
occlusor, often called adductor, impressions and a pair of large flabellate
divaricator impressions. In the dorsal valve the large “cardinal process” served
for the attachment of the divaricator, and a low median septum separated the
dendritic occlusor scars, which are rarely divisible into anterior and posterior
pairs.

Fig. 331.—Atrypa
reticularis. (Wenlock
Limestone.)

Fig. 332.—Interior of the

same, seen from the
dorsal side, showing
brachial spires. (After
Hall.)
In Leptaena (Fig. 334) the occlusor scars (a) in the ventral valve are narrow
and median, and are enclosed by a pair of flabelliform divaricator impressions
(d.v); in the dorsal valve two pairs of occlusor scars (a.a, p.a) are well marked,
and accessory posterior occlusor scars are traceable in some specimens. The
vascular sinuses (v.s) and genital areas are conspicuous in many species of
this and other genera.
 Fig. 333.—Productus giganteus. (After Woodward.) Carboniferous
Limestone. A, Interior of dorsal valve. B, Interior of ventral valve.
C, Transverse section of valves. D, Hinge line of A: a, occlusor
scars; d, divaricator scars; i, “reniform impressions”; ca, cardinal
process; h, hinge line; p, brachial prominence; s, cavity for spiral
arms; do, dorsal valve; ve, ventral valve.
In Waldheimia (Fig. 329) a sub-umbonal “peduncular muscle” scar (p) in the
ventral valve has before it a pair of “accessory divaricator” scars (a.d) flanked
by a pair of “ventral adjustor” (v.a) and a pair of “divaricator” impressions (d),
between which lie the two occlusor scars (a). In the dorsal valve anterior and
posterior pairs of occlusor scars (a.a, a.p) are visible.
The minute structure of the calcareous shell of the Testicardines is of
flattened fibrous prisms inclined at a very acute angle to the surfaces. In many
forms minute tubes more or less closely arranged pierce through the fibrous
shell-substance; but in some genera (Productus) they do not reach the outer
surface (see p. 468). Allied genera, however, differ much in the punctate or
impunctate character of the shell.
 Fig. 334.—Leptaena rhomboidalis. (Silurian.) A, External view of ventral
valve. B, Interior of ventral valve: a, occlusor scars; d, pseudo-deltidium;
d.v, divaricator scars; c.a, hinge area; t, teeth. C, Interior of dorsal valve:
a.a, anterior occlusor scars; p.a, posterior occlusor scars; c.a, hinge
area; c.p, cardinal process; d, chilidium; s, dental sockets; v.s, vascular
sinuses.
Synopsis of Families

I. Ecardines
Family. Lingulidae
Shell elongated, composed of alternating chitinous and calcareous layers, the
latter of which are perforated. Attached by a pedicle passing between apices of
valves.
Arms have no calcified supports.
(For muscles see Fig. 322.)
Range.—Lower Cambrian to Recent.
Principal Genera.—Lingula, Lingulella, Lingulepis.
Family. Obolidae
 Shell varies in shape. Ventral valve provided with pedicular groove or
foramen. Cardinal border thickened. No brachial supports. Shell composed of
alternating chitinous and calcareous layers.
(For muscles see p. 496.)
Range.—Lower Cambrian to Devonian.
Principal Genera.—Obolus, Obolella, Kutorgina, Linnarssonia,
Siphonotreta, Acrotreta, Neobolus.
Family. Discinidae
Shell rounded, valves more or less conical, fixed by pedicle passing through
slit or tubular foramen in ventral valve. No calcified brachial supports. Shell
structure chitino-calcareous.
Range.—Ordovician to Recent.
Principal Genera.—Discina, Orbiculoidea, Trematis.
Family. Craniidae
Shell calcareous, subcircular; fixed by surface of ventral valve; dorsal valve
the larger, depressed-conical. Shell structure punctate.
Four principal muscular scars in each valve, with central triangular
protuberance in ventral valve (see p. 476).
Range.—Ordovician to Recent.
Principal Genus.—Crania.
Family. Trimerellidae
Shell thick, calcareous, inequivalve; beak of ventral valve usually prominent;
rudimentary teeth maybe present; hinge area well developed, with pseudo-
deltidium. In interior of valves muscular platform, “crescent,” and sometimes
sub-umbonal chambers (see p. 494, Fig. 323).
Range.—Ordovician and Silurian; maximum in Wenlock.
Principal Genera.—Trimerella, Monomerella, Dinobolus, Rhinobolus.

II. Testicardines
Family. Productidae
Shell entirely free, or fixed by ventral valve or spines. Concavo-convex, more
or less covered with tubular spines. Hinge line straight. Hinge-teeth absent or
rudimentary.
Cardinal process prominent.
Reniform impressions in dorsal valve.
(For muscular impressions see p. 501, Fig. 333.)
Range.—Silurian to Permian. Genus Productus very characteristic of the
Carboniferous.
 Principal Genera.—Productus, Chonetes, Strophalosia, Proboscidella,
Aulosteges.
Family. Strophomenidae
Shell very variable in shape; concavo-convex, plano-convex, or biconvex;
hinge line usually straight; frequently with an area on each valve; foramen may
or may not be present. Shell structure near always punctate. Ventral valve
usually furnished with hinge-teeth; and dorsal valve with cardinal process.
Brachial supports completely absent or very rudimentary.
(For muscular impressions see p. 502, Fig. 334.)
Range.—Wholly Palaeozoic.
Principal Genera.—Orthis, with many sub-genera, Clitambonites,
Skenidium, Strophomena, Orthothetes, Leptaena, Stropheodonta,
Plectambonites.
Family. Koninckinidae
Shell plano-convex or concavo-convex. Brachial apparatus composed of two
lamellae spirally enrolled in the same plane, or in the form of depressed cones,
with the apices directed into the ventral valve.
Range.—Silurian to Lias.
Principal Genera.—Koninckina, Koninckella, Coelospira, Davidsonia.
Family. Spiriferidae
Shell biconvex. Brachial apparatus consisting essentially of two descending
calcareous lamellae which by spiral enrolment form a pair of laterally-directed
cones (Fig. 330).
Range.—Chiefly Palaeozoic, but a few forms pass up into the Lias.
Principal Genera.—Spirifera, Cyrtia, Uncites, Athyris, Merista.
Family. Atrypidae
Brachial apparatus consists of two descending calcareous lamellae which
bend outwards at the extremity of the crura and are coiled into two spiral cones,
the apices of which either converge towards each other (Glassia) or towards the
dorsal valve (Atrypa, Fig. 332), or diverge towards the dorsal valve (Dayia); shell
structure impunctate.
Range.—Ordovician to Trias.
Principal Genera.—Atrypa, Dayia, Glassia.
Family. Rhynchonellidae
Shell biconvex, hinge line usually curved.
Beak of ventral valve incurved, with foramen.
Calcareous brachial supports reduced to a pair of short curved crura.
The septa, dental and socket plates may be highly developed and divide up
the cavity of the shell into chambers (Stenochisma, Conchidium).
 Shell structure fibrous, rarely punctate; muscular impressions as in
Terebratulidae.
Range.—Ordovician to Recent: majority of the genera are Palaeozoic.
Principal Genera.—Rhynchonella (Fig. 327), Stenochisma, Stricklandia,
Conchidium.
Family. Terebratulidae
Shell structure punctate.
Arms supported by a calcareous loop, usually bent back on itself.
(For muscular impressions see p. 502, Figs. 328, 329.)
Beak of ventral valve perforated by foramen, furnished with deltidium.
Range.—Devonian to Recent; maximum development in Mesozoic times.
Principal Genera.—Terebratula, Terebratulina, Waldheimia, Terebratella,
Kingena, Magas, Centronella.
Family. Argiopidae
Large foramen for passage of pedicle. Marginal septa present in both valves.
Calcareous brachial loop follows margin of shell and is more or less fused with
the septa. Shell structure punctate.
Range.—Jurassic to Recent.
Principal Genera.—Argiope, Cistella.
Family. Stringocephalidae
Shell subcircular, punctate. Cardinal process highly developed, bifid. Brachial
apparatus composed of two calcareous free lamellae, prolonged at first
downwards, then bent back, upwards and outwards to run parallel to margin of
shell and to unite in front, thus constituting a wide loop.
Range.—Silurian and Devonian.
Sole Genus.—Stringocephalus.
Family. Thecidiidae
Shell usually fixed by beak of ventral valve, plano-convex. Sub-cardinal
apophysis in ventral valve for attachment of occlusors. Marginal septa in dorsal
valve. Calcareous brachial loop more or less fused with shell, and with
calcareous spicules of mantle. Shell structure: inner layer fibrous, outer layer
tubulated.
Range.—Carboniferous to Recent.
Principal Genera.—Thecidium, Oldhamina.

Stratigraphical Distribution of Brachiopoda

It is remarkable that some of the earliest types of Brachiopoda exist
generically unchanged at the present day. Such are Lingula, ranging from the
Cambrian; Discina and Crania, ranging from the Ordovician; and amongst the
hinged forms Terebratula from the Devonian, and Rhynchonella from the
Ordovician.
In the lowest Cambrian (Olenellus beds) the most important genera are
Linnarssonia and Kutorgina. The hinged forms appear in the Cambrian, being
represented by Orthis; but the majority in this formation belong to the
Ecardines. Lingula, Lingulella, and Obolella are characteristic.
In the Ordovician many new genera of the Testicardines make their
appearance, such as Strophomena, Leptaena, Atrypa, Rhynchonella,
Clitambonites, etc., but the extraordinary abundance and variety of Orthis is
most remarkable. The Ecardines are reinforced by such forms as Trematis and
Siphonotreta. It is, however, in the Silurian that the Testicardinate Brachiopoda
attain their maximum, for in addition to a great development of species
amongst the older forms, a host of new genera for the first time occur here
(Spirifera, Athyris, Conchidium, Stricklandia, Chonetes, Cyrtia, etc.); and the
Trimerellidae are especially characteristic of the Wenlock.
With the commencement of Devonian times many species and genera
become extinct, but new forms come in (Terebratula, Orthothetes, Productus,
etc.), and some genera are wholly confined to this formation (Uncites,
Stringocephalus). The Carboniferous is marked by the maximum development
of Productus and Spirifera; Orthothetes, Stenochisma, and Athyris are also
abundant, but there is a considerable extinction of the older genera and
species, and a great diminution in the number of individuals and species of
those that persist.
A further reduction occurs in the Permian, where the most important genera
are Productus, Strophalosia, and Stenochisma; but Aulosteges is a new form
peculiar to this period. In the Trias a new era commences; the principal
families and genera of the older rocks disappear entirely; a few spire-bearing
genera persist (Spiriferina, Athyris), and the genus Koninckina is restricted to
this formation.
The enormous development of species of the Terebratulidae and
Rhynchonellidae is the most noticeable feature in Jurassic times; and a few
ancient types linger on into the Lias (Spiriferina, Suessia, a sub-genus of
Spirifera); Koninckella here occurs.
The Cretaceous Brachiopoda are closely allied to the Jurassic; Magas and
Lyra are peculiar to the period, and the Terebratulidae and Rhynchonellidae
are very abundant, together with the Ecardinate genus Crania.
With the commencement of Tertiary times the Brachiopoda have lost their
geological importance, and have dwindled down into an insignificant
proportion of the whole Invertebrate fauna.

The distribution of the Brachiopoda in past time is shown in the following

table:—
Palaeozoic Mesozoic
C
a
r
O b C
r o r
C d S D n J e T
a o i e i P u t e
m v l v f e r a r R
b i u o e r T a c t e
r c r n r m r s e i c
i i i i o i i s o a e
a a a a u a a i u r n
ECARDINES n n n n s n s c s y t
Lingulidae Lingula ___ ___ ___ ___ ___ ___ ___ ___ ___ ___ ___
Lingulella ___
Obolidae Obolus ___ ___
Obolella ___ ___
Kutorgina ___ ___
Linnarssonia ___
Trematis ___ ___
Siphonotreta ___ ___
Acrotreta ___
Discinidae Discina ___ ___ ___ ___ ___ ___ ___ ___ ___ ___
Craniidae Crania ___ ___ ___ ___ ___ ___ ___ ___ ___ ___
Trimerellidae Trimerella ___
Dinobolus ___
TESTICARDINES
Productidae Productus ___ ___ ___
Chonetes ___ ___ ___
Strophalosia ___ ___ ___
Strophomenidae Orthis ___ ___ ___ ___ ___
Skenidium ___ ___
Clitambonites ___
Strophomena ___ ___
Stropheodonta ___ ___ ___
Leptaena ___ ___ ___ ___
Orthothetes ___ ___ ___
Davidsonia ___
Koninckinidae Koninckina ___
 Koninckella ___
Spiriferidae Spirifera ___ ___ ___ ___
Spiriferina ___ ___ ___ ___ ___
Cyrtia ___ ___ ___
Syringothyris ___
Uncites ___
Athyris ___ ___ ___ ___ ___
Merista ___ ___
Retzia ___ ___ ___ ___ ___
Atrypidae Atrypa ___ ___ ___ ___ ___ ___
Dayia ___
Coelospira ___
Rhynchonellidae Rhynchonella ___ ___ ___ ___ ___ ___ ___ ___ ___ ___
Stenochisma ___ ___ ___
Stricklandia ___
Conchidium ___ ___
Terebratulidae Terebratula ___ ___ ___ ___ ___ ___ ___ ___
Terebratulina ___ ___ ___ ___
Waldheimia ___ ___ ___ ___
Terebratella ___ ___ ___ ___
Kingena ___ ___
Magas ___
Centronella ___ ___ ___
Argiopidae Argiope ___ ___ ___ ___
Cistella ___ ___ ___ ___
Stringocephalidae Stringocephalus ___
Thecidiidae Thecidium ___ ___ ___ ___ ___
Oldhamina ___

PHYLOGENY AND ONTOGENY

Wherever successive stages in the life history of an individual resemble in

important anatomical features the adult individuals of other species occurring
in successive members of a stratigraphical series, the development of the
individual may be regarded as an epitome of the development of the species;
it also generally throws light on the origin and relationships of allied genera
and families.
In the case of the fossil Brachiopoda comparatively little work has yet been
done in tracing their ontogeny or phylogeny, though the abundance, variety,
and excellent state of preservation of the extinct species offer a promising field
for investigation. It is to Dr. C. E. Beecher and other recent American
palaeontologists that we owe our advance in this branch of the subject.
In the first place, in about forty genera, representing nearly all the leading
families of the group, the important fact has been established of the presence
of a common form of embryonic shell, termed the “protegulum,” which is
“semicircular or semielliptical in shape with a straight or arcuate hinge line and
no hinge area” (Beecher).[431] Its minute size and delicate texture cause its
preservation to be rare, but its impression is not uncommonly left on the beak
of the adult shell.
The main features of this embryonic shell are exhibited in the adult Lower
Cambrian Brachiopod Obolus (Kutorgina) labradoricus (Billings); the sub-equal
semielliptical valves have lines of growth running concentrically and parallel to
the margin of the shell, and ending abruptly against the straight hinge line; and
this indicates that there has been no change in the outline and proportions of
the shell during its stages of growth, but only a general increase in size. It is
very significant that we have here a mature type possessing the common
embryonic characters of a host of widely separated genera, and we may
therefore regard it as the most primitive form known.
Many genera pass through this so-called “Paterina” stage either in the case
of both their valves, or more generally in the case of the dorsal valve only; but
modifications in the form of the protegulum arise, which are due to the
influence of accelerated growth, by which features belonging to later stages
become impressed on the early embryonic shell. The most variable and
specialised valve—the ventral or pedicle valve naturally exhibits the effect of
this influence first and to the greatest extent. The Palaeozoic adult forms of
many species represent various pre-adult stages of the Mesozoic, Tertiary, and
Recent species, as is especially well shown in the genera Orbiculoidea and
Discinisca.
In the Strophomenoid shells the protegulum in the dorsal valve is usually
normal, but in the ventral valve abbreviation of the hinge and curvature of the
hinge line are produced by acceleration of the “Discinoid stage” in which a
pedicle notch is present.
No marked variation has yet been noticed in the spire-bearing, or
Terebratuloid, or Rhynchonelloid genera.
The form of the shell and the amount of difference in shape and size of the
valves seem to be largely due to the length of the pedicle and its inclination to
the axis of the body, as evidenced by the development of Terebratulina. A
series showing progressive dissimilarity of the two valves arising from these
causes can be traced from Lingula to Crania. The greater alteration that takes
place in the ventral valve appears to be due to its position as lower and
attached valve. If the pedicle is short a transversely-expanded shell with long
hinge line results when the plane of the valves is vertical or ascending, but
when the latter is horizontal a Discinoid form is found. This mode of
attachment is often accompanied by a more or less plainly developed radial
symmetry. Shells with long pedicles, on the other hand, are usually longer than
wide.
 The character of the pedicle-opening is of great significance from an
evolutional and classificatory point of view, for the successive stages through
which it passes in embryonic growth are chronologically paralleled by different
genera, and are likewise accompanied by the successive acquisition of other
important anatomical characters, as has been shown by Beecher and others.
The first and simplest type of pedicle opening is in shells with a posterior
gaping of the valves, where the pedicle protrudes freely between them in a line
with the axis, and the opening is shared by both valves, though generally to a
greater extent by the ventral valve. Paterina (= Obolus labradoricus) and
Lingula furnish examples of this type. In the second type the pedicle opening
is restricted to the ventral valve, and the direction of the pedicle makes a right
angle with the plane of the valves; in the lower forms the pedicle lies in a slit or
sinus (Trematidae), but by further specialisation it becomes enclosed by shell
growth so as to lie within the periphery, and finally becomes sub-central in
some genera (Discinidae). The third type shows the pedicle opening confined
to the ventral valve and sub-marginal. A pseudo-deltidium may preserve the
original opening (Clitambonites); or this shelly plate may become worn away
or reabsorbed in the adult so that the deltidial fissure through which the
pedicle passes remains quite open (Orthidae). In the fourth type the incipient
stage marks a return to the simple conditions of the first type; but ultimately a
pair of deltidial plates develop, and may completely limit the pedicle opening
below. Examples of this type are Spirifera and Rhynchonella. By means of
these four types the Brachiopods have been divided into four Orders: the
Atremata (type i.); the Neotremata (type ii.); the Protremata (type iii.); and the
Telotremata (type iv.).
The Telotremata were the last to appear, but the four types of pedicle-
opening with the various forms of calcareous brachial apparatus were in
existence in the Bala period of the Ordovician.
As Paterina is the most primitive form of all, we may place it at the root of
the phylogenetic tree. From it sprang the Atremata, which gave off the
Neotremata and Protremata; the most primitive Neotremata seem to be the
Trematidae, while the connecting link between the Protremata and Atremata is
furnished by the Kutorginidae. From the genus Conchidium and its allies we
may see how the Rhynchonellidae ushered in the Telotremata as an offshoot
from the Protremata. The Telotremata subsequently gave off two main
branches, which became specialised with the loop-bearing and spire-bearing
forms respectively.
The evolution and mutual relationships of genera have been indicated with
much probability by Hall, Clarke, and others. The Obolelloid type may be
connected with the Linguloid by means of Lingulella and Linyulepis, while in
Lingula itself we find the point of divergence for the ancestors of Trimerella,
and for a line of variation culminating in Dignomia. The Palaeozoic
Rhynchonelloids branched off at an early period from the same stock as
Orthis, and are connecting links between this genus and Mesozoic
Rhynchonellae; and a whole series of genera exhibit intermediate stages of
structure between the Rhynchonelloid and Pentameroid groups. The
Terebratuloids can be traced back to the primitive type Renssoellaria; and
amongst spire-bearing forms, the protean genus Spirifera can be split up into
groups of species which diverge along lines tending to forms no longer
congeneric. When we come to deal with specific differences we find frequently
such a host of intermediate varieties that the separation of many species, as in
the case of Mesozoic Terebratulae, is to a large extent arbitrary and artificial.