(eBook PDF) Computer Systems 5th Edition download pdf
(eBook PDF) Computer Systems 5th Edition download pdf
com
https://ebookluna.com/product/ebook-pdf-computer-
systems-5th-edition/
OR CLICK BUTTON
DOWNLOAD NOW
https://ebookluna.com/product/computer-systems-5th-edition-ebook-pdf/
ebookluna.com
https://ebookluna.com/product/ebook-pdf-automation-production-systems-
and-computer-integrated-manufacturing-5th-edition/
ebookluna.com
https://ebookluna.com/product/ebook-pdf-computer-systems-a-
programmers-perspective-3nd-edition/
ebookluna.com
ebookluna.com
https://ebookluna.com/download/data-prefetching-techniques-in-
computer-systems-ebook-pdf/
ebookluna.com
https://ebookluna.com/download/progress-in-heterocyclic-chemistry-
ebook-pdf/
ebookluna.com
https://ebookluna.com/product/ebook-pdf-computer-architecture-a-
quantitative-approach-5th-edition/
ebookluna.com
https://ebookluna.com/product/ebook-pdf-translational-medicine-in-cns-
drug-development-volume-29/
ebookluna.com
Problems
3. Information Representation
3.1 Unsigned Binary Representation
Binary Storage
Integers
Base Conversions
Range for Unsigned Integers
Unsigned Addition
The Carry Bit
3.2 Two’s Complement Binary Representation
Two’s Complement Range
Base Conversions
The Number Line
The Overflow Bit
The Negative and Zero Bits
3.3 Operations in Binary
Logical Operators
Register Transfer Language
Arithmetic Operators
Rotate Operators
3.4 Hexadecimal and Character Representations
Hexadecimal
Base Conversions
ASCII Characters
Unicode Characters
3.5 Floating-Point Representation
Binary Fractions
Excess Representations
The Hidden Bit
Special Values
The IEEE 754 Floating-Point Standard
8
3.6 Models
Chapter Summary
Exercises
Problems
4. Computer Architecture
4.1 Hardware
Central Processing Unit (CPU)
Main Memory
Input/Output Devices
Data and Control
Instruction Format
4.2 Direct Addressing
The Stop Instruction
The Load Word Instruction
The Store Word Instruction
The Add Instruction
The Subtract Instruction
The And and Or Instructions
The Invert and Negate Instructions
The Load Byte and Store Byte Instructions
The Input and Output Devices
Big Endian Versus Little Endian
4.3 von Neumann Machines
The von Neumann Execution Cycle
A Character Output Program
von Neumann Bugs
A Character Input Program
Converting Decimal to ASCII
A Self-Modifying Program
4.4 Programming at Level ISA3
Read-Only Memory
The Pep/9 Operating System
Using the Pep/9 System
Chapter Summary
9
Exercises
Problems
Level 5 Assembly
5. Assembly Language
5.1 Assemblers
Instruction Mnemonics
Pseudo-Operations
The .ASCII and .END Pseudo-ops
Assemblers
The .BLOCK Pseudo-op
The .WORD and .BYTE Pseudo-ops
Using the Pep/9 Assembler
Cross Assemblers
5.2 Immediate Addressing and the Trap Instructions
Immediate Addressing
The DECI, DECO, and BR Instructions
The STRO Instruction
Interpreting Bit Patterns: The HEXO Instruction
Disassemblers
5.3 Symbols
A Program with Symbols
A von Neumann Illustration
5.4 Translating from Level HOL6
The Printf() Function
Variables and Types
Global Variables and Assignment Statements
Type Compatibility
Pep/9 Symbol Tracer
The Shift and Rotate Instructions
Constants and .EQUATE
Placement of Instructions and Data
Chapter Summary
10
Exercises
Problems
11
Translating Structures
Translating Linked Data Structures
Chapter Summary
Exercises
Problems
12
Exercises
Problems
8. Process Management
8.1 Loaders
The Pep/9 Operating System
The Pep/9 Loader
Program Termination
8.2 Traps
The Trap Mechanism
The RETTR Instruction
The Trap Handlers
Trap Addressing Mode Assertion
Trap Operand Address Computation
The No-Operation Trap Handlers
The DECI Trap Handler
The DECO Trap Handler
The HEXO and STRO Trap Handlers and Operating System
Vectors
8.3 Concurrent Processes
Asynchronous Interrupts
Processes in the Operating System
Multiprocessing
A Concurrent Processing Program
Critical Sections
A First Attempt at Mutual Exclusion
A Second Attempt at Mutual Exclusions
Peterson’s Algorithm for Mutual Exclusion
Semaphores
Critical Sections with Semaphores
8.4 Deadlocks
Resource Allocation Graphs
13
Deadlock Policy
Chapter Summary
Exercises
Problems
9. Storage Management
9.1 Memory Allocation
Uniprogramming
Fixed-Partition Multiprogramming
Logical Addresses
Variable-Partition Multiprogramming
Paging
9.2 Virtual Memory
Large Program Behavior
Virtual Memory
Demand Paging
Page Replacement
Page-Replacement Algorithms
9.3 File Management
Disk Drives
File Abstraction
Allocation Techniques
9.4 Error-Detecting and Error-Correcting Codes
Error-Detecting Codes
Code Requirements
Single-Error-Correcting Codes
9.5 RAID Storage Systems
RAID Level 0: Nonredundant Striped
RAID Level 1: Mirrored
RAID Levels 01 and 10: Striped and Mirrored
RAID Level 2: Memory-Style ECC
RAID Level 3: Bit-Interleaved Parity
RAID Level 4: Block-Interleaved Parity
RAID Level 5: Block-Interleaved Distributed Parity
Chapter Summary
14
Exercises
15
Exercises
Level 2 Microcode
16
The Store Word Direct Instruction
The Add Immediate Instruction
The Load Word Indirect Instruction
The Arithmetic Shift Right Instruction
The CPU Control Section
12.2 Performance
The Data Bus Width and Memory Alignment
Memory Alignment
The Definition of an n-Bit Computer
Cache Memories
The System Performance Equation
RISC Versus CISC
12.3 The MIPS Machine
The Register Set
The Addressing Modes
The Instruction Set
MIPS Computer Organization
Pipelining
12.4 Conclusion
Simplifications in the Model
The Big Picture
Chapter Summary
Exercises
Problems
Appendix
Solutions to Selected Exercises
Index
17
Preface
The fifth edition of Computer Systems offers a clear, detailed, step-by-step
exposition of the central ideas in computer organization, assembly
language, and computer architecture. The book is based in large part on a
virtual computer, Pep/9, which is designed to teach the basic concepts of
the classic von Neumann machine. The strength of this approach is that the
central concepts of computer science are taught without getting entangled
in the many irrelevant details that often accompany such courses. This
approach also provides a foundation that encourages students to think
about the underlying themes of computer science. Breadth is achieved by
emphasizing computer science topics that are related to, but not usually
included in, the treatment of hardware and its associated software.
Summary of Contents
Computers operate at several levels of abstraction; programming at a high
level of abstraction is only part of the story. This book presents a unified
concept of computer systems based on the level structure of FIGURE
P.1 .
The book is divided into seven parts, corresponding to the seven levels
of Figure P.1:
18
The text generally presents the levels top-down, from the highest to the
lowest. Level ISA3 is discussed before Level Asmb5, and Level LG1 is
discussed before Level Mc2 for pedagogical reasons. In these two
instances, it is more natural to revert temporarily to a bottom-up approach
so that the building blocks of the lower level will be in hand for
construction of the higher level.
Level App7
Level App7 is a single chapter on application programs. It presents the
idea of levels of abstraction and binary information and establishes the
framework for the remainder of the book. A few concepts of relational
databases are presented as an example of a typical computer application.
Level HOL6
19
Level HOL6 consists of one chapter, which reviews the C programming
language. The chapter assumes that the student has experience in some
imperative language, such as Java or Python—not necessarily C. The
instructor can readily translate the C examples to other common Level
HOL6 languages if necessary.
This chapter emphasizes the C memory model, including global versus
local variables, functions with parameters, and dynamically allocated
variables. The topic of recursion is treated because it depends on the
mechanism of memory allocation on the run-time stack. A fairly detailed
explanation is given on the details of the memory allocation process for
function calls, as this mechanism is revisited at a lower level of abstraction
later in the book.
Level ISA3
Level ISA3 is the instruction set architecture level. Its two chapters
describe Pep/9, a virtual computer designed to illustrate computer
concepts. Pep/9 is a small complex instruction set computer (CISC); a von
Neumann machine. The central processing unit (CPU) contains an
accumulator, an index register, a program counter, a stack pointer, and an
instruction register. It has eight addressing modes: immediate, direct,
indirect, stack-relative, stack-relative deferred, indexed, stack-indexed, and
stack-deferred indexed. The Pep/9 operating system, in simulated read-
only memory (ROM), can load and execute programs in hexadecimal
format from students’ text files. Students run short programs on the Pep/9
simulator and learn that executing a store instruction to ROM does not
change the memory value.
Students learn the fundamentals of information representation and
computer organization at the bit level. Because a central theme of this
book is the relationship of the levels to one another, the Pep/9 chapters
show the relationship between the ASCII representation (Level ISA3) and
C variables of type char (Level HOL6). They also show the relationship
between two’s complement representation (Level ISA3) and C variables of
type int (Level HOL6).
Level Asmb5
Level Asmb5 is the assembly level. The text presents the concept of the
assembler as a translator between two levels—assembly and machine. It
introduces Level Asmb5 symbols and the symbol table.
The unified approach really comes into play here. Chapters 5 and 6
present the compiler as a translator from a high-order language to
20
assembly language. Previously, students learned a specific Level HOL6
language, C, and a specific von Neumann machine, Pep/9. These chapters
continue the theme of relationships between the levels by showing the
correspondence between (a) assignment statements at Level HOL6 and
load/store instructions at Level Asmb5, (b) loops and if statements at Level
HOL6 and branching instructions at Level Asmb5, (c) arrays at Level
HOL6 and indexed addressing at Level Asmb5, (d) procedure calls at
Level HOL6 and the run-time stack at Level Asmb5, (e) function and
procedure parameters at Level HOL6 and stack-relative addressing at
Level Asmb5, (f) switch statements at Level HOL6 and jump tables at
Level Asmb5, and (g) pointers at Level HOL6 and addresses at Level
Asmb5.
The beauty of the unified approach is that the text can implement the
examples from the C chapter at this lower level. For example, the run-time
stack illustrated in the recursive examples of Chapter 2 corresponds
directly to the hardware stack in Pep/9 main memory. Students gain an
understanding of the compilation process by translating manually between
the two levels.
This approach provides a natural setting for the discussion of central
issues in computer science. For example, the book presents structured
programming at Level HOL6 versus the possibility of unstructured
programming at Level Asmb5. It discusses the goto controversy and the
structured programming/efficiency tradeoff, giving concrete examples
from languages at the two levels.
Chapter 7, “Language Translation Principles,” introduces students to
computer science theory. Now that students know intuitively how to
translate from a high-level language to assembly language, we pose the
fundamental question underlying all of computing: What can be
automated? The theory naturally fits in here because students now know
what a compiler (an automated translator) must do. They learn about
parsing and finite-state machines—deterministic and nondeterministic—in
the context of recognizing C and Pep/9 assembly language tokens. This
chapter includes an automatic translator between two small languages,
which illustrates lexical analysis, parsing, and code generation. The lexical
analyzer is an implementation of a finite-state machine. What could be a
more natural setting for the theory?
Level OS4
Level OS4 consists of two chapters on operating systems. Chapter 8 is a
description of process management. Two sections, one on loaders and
21
another on trap handlers, illustrate the concepts with the Pep/9 operating
system. Seven instructions have unimplemented opcodes that generate
software traps. The operating system stores the process control block of
the user’s running process on the system stack, and the interrupt service
routine interprets the instruction. The classic state transition diagram for
running and waiting processes in an operating system is thus reinforced
with a specific implementation of a suspended process. The chapter
concludes with a description of concurrent processes and deadlocks.
Chapter 9 describes storage management, both main memory and disk
memory.
Level LG1
Level LG1 uses two chapters to present combinational and sequential
circuits. Chapter 10 emphasizes the importance of the mathematical
foundation of computer science by starting with the axioms of Boolean
algebra. It shows the relationship between Boolean algebra and logic gates
and then describes some common logic devices, including a complete
logic design of the Pep/9 arithmetic logic unit (ALU). Chapter 11
illustrates the fundamental concept of a finite-state machine through the
state transition diagrams of sequential circuits. It concludes with a
description of common computer subsystems such as bidirectional buses,
memory chips, and two-port memory banks.
Level Mc2
Chapter 12 describes the microprogrammed control section of the Pep/9
CPU. It gives the control sequences for a few sample instructions and
addressing modes and provides a large set of exercises for the others. It
also presents concepts of load/store architectures, contrasting the MIPS
reduced instruction set computer (RISC) machine with the Pep/9 CISC
machine. It concludes with performance issues by describing cache
memories, pipelining, dynamic branch prediction, and superscalar
machines.
Use in a Course
This book offers such broad coverage that instructors may wish to omit
some of the material when designing a course. I use Chapters 1–7 in a
computer systems course and Chapters 10–12 in a computer organization
course.
In the book, Chapters 1–5 must be covered sequentially. Chapters 6
22
(“Compiling to the Assembly Level”) and 7 (“Language Translation
Principles”) can be covered in either order. I often skip ahead to Chapter 7
to initiate a large software project, writing an assembler for a subset of
Pep/9 assembly language, so students will have sufficient time to complete
it during the semester. Chapter 11 (“Sequential Circuits”) is obviously
dependent on Chapter 10 (“Combinational Circuits”), but neither depends
on Chapter 9 (“Storage Management”), which may be omitted. FIGURE
P.2 , a chapter dependency graph, summarizes the possible chapter
omissions.
23
allocation from the heap uses malloc instead of new.
› Sidebars—In this edition, the brief biographies formerly found in each
chapter have been replaced with a different set of sidebars. Each
sidebar is a real-world example of the concepts described in that
chapter. As most of the chapters describe the Pep/9 virtual machine,
the sidebars for those chapters show some corresponding
implementations for the Intel x86 architecture. The new sidebars give
a consistent running example of this architecture so students have a
better idea of how the concepts of the virtual machine relate to real-
world implementations.
› New and expanded topics—Chapter 1 now emphasizes how binary
information is quantified in both space and time by presenting the
performance equation and the concept of bandwidth. QR codes and
color displays are detailed examples of these concepts. Chapter 3
describes Unicode as well as UTF-32 and UTF-8 encoding. Chapter 4
discusses big-endian and little-endian order. Chapter 7 uses Java
instead of C++ as the implementation language for the example
translator. The microcode examples in Chapter 12 use the new
UnitPre and UnitPost feature of the Pep/9 CPU simulator. The two-
byte data bus now has simulator support, and that topic leads to an
extended discussion of memory alignment issues and the new .ALIGN
assembler directive.
Pep/8, the virtual machine for the two previous editions, has been
superseded by Pep/9. As the instruction sets of the two machines are
different, Pep/8 source and object programs are not compatible with those
of Pep/9. Only a few instructions are affected, however, so much remains
the same, including the set of eight addressing modes. The changes for
Pep/9 include the following:
24
data and the return mechanism for flow of control. Combining these
two different concepts into one statement can be confusing during the
learning process. Pep/9 now requires students to explicitly deallocate
locals with the ADDSP statement. An added stylistic advantage is that
the explicit ADDSP to deallocate locals at the end of a function
corresponds directly to the SUBSP at the beginning of the function to
allocate locals.
› Memory-mapped I/O—Of all the instructions in the Pep/8 instruction
set, the most unrealistic are CHARI and CHARO for character input
and output. Most real computer systems map input and output devices
to main memory, which is now the design of Pep/9. In the new
instruction set, there are no native input and output instructions.
Instead, the Pep/8 instruction
CHARI alpha,ad
25
Random documents with unrelated
content Scribd suggests to you:
Synagris.—Body rather elongate, covered with ciliated scales of
moderate size. Cleft of the mouth horizontal, with the jaws equal in
length anteriorly. One continuous dorsal, with feeble spines; dorsal
10/9, anal 3/7. Caudal deeply forked. Teeth villiform, with canines at
least in the upper jaw. Infraorbital not armed; præoperculum without,
or with a very indistinct serrature. Cheek with three series of scales.
Branchiostegals six.
Marine fishes of small size; about twenty species are known from
the tropical parts of the Indo-Pacific. Pentapus, Chætopterus, and
Aphareus are allied genera from the same area.
Maena.—Body oblong, compressed, covered with ciliated scales
of moderate size. Mouth very protractile, the intermaxillary pedicles
extending backwards to the occiput. Teeth villiform; minute teeth on
the vomer. One dorsal, scaleless, with feeble spines. D. 11/11, A. 3/9.
Caudal fin forked. Præoperculum without serrature. Branchiostegals
six.
Small fishes from the Mediterranean, known to the ancients;
valueless as food. Three species.
Smaris.—Body oblong or cylindrical, covered with rather small
ciliated scales. Mouth very protractile, the intermaxillary pedicles
extending backwards to the occiput. Teeth villiform. Palate toothless.
One dorsal, scaleless, with eleven or more very feeble spines; anal
with three. Caudal fin forked. Præoperculum without serrature.
Branchiostegals six.
Small fishes from the Mediterranean. Six species.
Cæsio.—Body oblong, covered with ciliated scales of moderate
size. Cleft of the mouth more or less oblique, with the jaws equal in
length anteriorly, or with the lower somewhat projecting. Teeth
villiform; palate generally toothless. One dorsal, with from nine to
thirteen very feeble spines, with the anterior part highest, and the
posterior covered with minute scales. Caudal fin deeply forked.
Præoperculum without, or with minute, serrature.
Small fishes from the Indo-Pacific. Twelve species.
Erythrichthys.—Body elongate, covered with small ciliated
scales. Mouth very protractile, the pedicles of the intermaxillary
extending to the occiput. Dentition quite rudimentary or entirely absent
Two dorsal fins connected by a series of very feeble spines; also the
anterior spines are feeble. Præoperculum not serrated.
Fig. 161.—Erythrichthys nitidus.
Second Family—Squamipinnes.
Body compressed and elevated, covered with scales, either finely
ctenoid or smooth. Lateral line continuous, not continued over the
caudal fin. Mouth in front of the snout, generally small, with lateral
cleft. Eye lateral, of moderate size. Six or seven branchiostegals.
Teeth villiform or setiform, in bands, without canines or incisors.
Dorsal fin consisting of a spinous and soft portion of nearly equal
development; anal with three or four spines, similarly developed as
the soft dorsal, both being many-rayed. The vertical fins more or less
densely covered with small scales. The lower rays of the pectoral fin
branched, not enlarged; ventrals thoracic, with one spine and five
soft rays. Stomach coecal.
The typical forms of this family are readily recognised by the form
of their body, and by a peculiarity from which they derive their name
Squamipinnes; the soft, and frequently also the spinous part of their
dorsal and anal fins are so thickly covered with scales that the
boundary between fins and body is entirely obliterated. The majority
are inhabitants of the tropical seas, and abound chiefly in the
neighbourhood of coral-reefs. The beauty and singularity of
distribution of the colours of some of the genera, as Chætodon,
Heniochus, Holacanthus, is scarcely surpassed by any other group
of fishes. They remain within small dimensions, and comparatively
few are used as food. They are carnivorous, feeding on small
invertebrates. Only a few species enter brackish water.
Extinct representatives of this family are not scarce at Monte
Bolca and in other tertiary formations. All, at least those admitting of
definite determination, belong to existing genera, viz. Holacanthus,
Pomacanthus, Ephippium, Scatophagus. Very singular is the
occurrence of Toxotes in the Monte Bolca strata.
The following genera have no teeth on the palate:—
Chætodon.—One dorsal fin, without any notch in its upper
margin, and with the soft and spinous portions similarly developed;
none of the spines elongate. Snout short or of moderate length.
Præoperculum without, or with a fine, serration, and without spine at
the angle. Scales generally large or of moderate size.
Fig. 167.—Chætodon ephippium.
Seventy species are known from the tropical parts of the Atlantic
and Indo-Pacific, nearly all being beautifully ornamented with bands
or spots. Of the ornamental markings a dark or bicoloured band,
passing through the eye and ascending towards the back, is very
generally found in these fishes; it frequently occurs again in other
marine Acanthopterygians, in which it is not rarely a sign of the
immature condition of the individual. The Chætodonts are most
numerous in the neighbourhood of the coral-reefs of the Indo-Pacific,
the species figured (C. ephippium) being as common in the East
Indian Archipelago as in Polynesia, like many others of its
congeners.
Chelmo differs from Chætodon only in having the snout produced
into a more or less long tube.
Fig. 168.—Chelmo marginalis, from the coast of Australia.
Only four species are known, locally distributed in the tropical
seas. Ch. rostratus, the oldest species known, is said to have the
instinct of throwing a drop of water from its bill so as to light upon
any insect resting on a leaf, and thus make it fall, that it may instantly
dart upon it. This statement is erroneous, and probably rests upon
the mistaken notion that the long bill is especially adapted for this
manœuvre, which, indeed, is practised by another fish of this family
(Toxotes). The long slender bill of Chelmo (which is a true saltwater
fish) rather enables it to draw from holes and crevices animals which
otherwise could not be reached by it.
Heniochus.—One dorsal, with from eleven to thirteen spines, the
fourth of which is more or less elongate and filiform. Snout rather short
or of moderate length. Præoperculum without spine. Scales of
moderate size.
Four species are known from the tropical Indo-Pacific. H.
macrolepidotus is one of the most common fishes of that area; the
species figured (H. varius) retains in a conspicuous manner horn-like
protuberances on bones of the head, with which the young of all the
species of this genus seem to be armed.
Two or three species are known from the warmer parts of the
Atlantic and Indian Oceans. The Atlantic species (E. faber) shows
the remarkable peculiarity that in old specimens (12 and more inches
long) the occipital crest, and sometimes some of the anterior neural
and hæmal spines are enormously enlarged into a globular bony
mass. This can hardly be regarded as a pathological change of the
bone, as it has been found in all old specimens, without exception.
Drepane is allied to Ephippus, but has very long falciform
pectoral fins. The single species D. punctata is common in the Indian
Ocean and on the coasts of Australia. Hypsinotus, from Japan,
appears to inhabit a greater depth than the other Squamipinnes.
Scorpis and Atypichthys are genera distinguished from the
preceding by the presence of vomerine teeth. They belong to the
coast-fauna of Australia, New Zealand, and Chili.
Toxotes.—Body short, compressed, covered with scales of
moderate size. Snout pointed, with a wide lateral mouth and projecting
lower jaw. One dorsal, with five strong spines situated on the posterior
part of the back; the soft portion and the anal fin scaly, the latter with
three spines. Villiform teeth in the jaws, on the vomer and palatine
bones. Scales of moderate size, cycloid.
Two species are known from the East Indies, one (T. jaculator),
which is the more common, ranging to the north coast of Australia. It
has received its name from its habit of throwing a drop of water at an
insect which it perceives close to the surface, in order to make it fall
into it. The Malays, who call it “Ikan sumpit,” keep it in a bowl, in
order to witness this singular habit, which it continues even in
captivity.
Third Family—Mullidæ.
Body rather low and slightly compressed, covered with large thin
scales, without or with an extremely fine serrature. Two long erectile
barbels are suspended from the hyoid, and are received between the
rami of the lower jaw and opercles. Lateral line continuous. Mouth in
front of the snout, with the cleft lateral and rather short; teeth very
feeble. Eye lateral, of moderate size. Two short dorsal fins remote
from each other, the first with feeble spines; anal similar to the
second dorsal. Ventrals with one spine and five rays. Pectorals short.
Branchiostegals four; stomach siphonal.
The “Red Mullets” form a very natural family, which, on account
of slight modifications of the dentition, has been divided into several
sub-genera—Upeneoides, Upeneichthys, Mullus, Mulloides, and
Upeneus. They are marine fishes, but many species enter brackish
water to feed on the animalcules abounding in the flora of brack-
water. About forty different species are known chiefly from tropical
seas, the European species (M. barbatus, see p. 43, Fig. 7),
extending far northwards into the temperate zone. None attain to a
large size, specimens of from two to three lbs. being not common,
but all are highly esteemed as food.
The most celebrated is the European species (of which there is
one only, M. surmuletus being probably the female). The ancient
Romans called it Mullus, the Greeks τριγλη. The Romans priced it
above any other fish; they sought for large specimens far and wide,
and paid ruinous prices for them.
“Mullus tibi quatuor emptus
Librarum, cœnæ pompa caputque fuit,
Exclamare libet, non est hic improbe, non est
Piscis: homo est; hominem, Calliodore, voras.”
Martial, x. 31.
Fourth Family—Sparidæ.
Body compressed, oblong, covered with scales, the serrature of
which is very minute, and sometimes altogether absent. Mouth in
front of the snout, with the cleft lateral. Eye lateral, of moderate size.
Either cutting teeth in front of the jaws, or molar teeth on the side;
palate generally toothless. One dorsal fin, formed by a spinous and
soft portion of nearly equal development. Anal fin with three spines.
The lower rays of the pectoral fin are generally branched, but in one
genus simple. Ventrals thoracic, with one spine and five rays.
The “Sea-breams” are recognised chiefly by their dentition, which
is more specialised than in the preceding families, and by which the
groups, into which this family has been divided, are characterised.
They are inhabitants of the shores of all the tropical and temperate
seas. Their coloration is very plain. They do not attain to a large size,
but the majority are used as food.
The extinct forms found hitherto are rather numerous; the oldest
come from the cretaceous formation of Mount Lebanon; some
belong to living genera, as Sargus, Pagellus; of others from Eocene
and Miocene formations no living representative is known—
Sparnodus, Sargodon, Capitodus, Soricidens, Asima.
First Group—Cantharina.—More or less broad cutting,
sometimes lobate, teeth in front of the jaws; no molars or vomerine
teeth; the lower pectoral rays are branched. Partly herbivorous,
partly carnivorous. The genera belonging to this group are:—
Cantharus from the European and South African coasts, of which
one species (C. lineatus), is common on the coasts of Great Britain,
and locally known by the names “Old Wife,” “Black Sea-bream;” Box,
Scatharus, and Oblata from the Mediterranean and neighbouring
parts of the Atlantic; Crenidens and Tripterodon from the Indian
Ocean; Pachymetopon, Dipterodon, and Gymnocrotaphus from the
Cape of Good Hope; Girella and Tephræops from Chinese,
Japanese, and Australian Seas; Doydixodon from the Galapagos
Islands and the coasts of Peru.
Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.
ebookluna.com