Download the full version of the ebook now at ebooksecure.

com

(eBook PDF) Digital Control System Analysis &

Design 4th Edition

https://ebooksecure.com/product/ebook-pdf-digital-
control-system-analysis-design-4th-edition/

Explore and download more ebook at https://ebooksecure.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

(eBook PDF) Power System Analysis and Design 6th Edition

https://ebooksecure.com/product/ebook-pdf-power-system-analysis-and-
design-6th-edition/

ebooksecure.com

(eBook PDF) Digital Logic Circuit Analysis and Design 2nd

Edition

https://ebooksecure.com/product/ebook-pdf-digital-logic-circuit-
analysis-and-design-2nd-edition/

ebooksecure.com

(eBook PDF) Digital Design: Principles and Practices 4th

https://ebooksecure.com/product/ebook-pdf-digital-design-principles-
and-practices-4th/

ebooksecure.com

Flow Assurance. Volume II 1st Edition - eBook PDF

https://ebooksecure.com/download/flow-assurance-volume-ii-ebook-pdf/

ebooksecure.com
(eBook PDF) Elemental Geosystems, 7th Edition

https://ebooksecure.com/product/ebook-pdf-elemental-geosystems-7th-
edition/

ebooksecure.com

(Original PDF) The Anterior Cruciate Ligament

Reconstruction and Basic Science

https://ebooksecure.com/product/original-pdf-the-anterior-cruciate-
ligament-reconstruction-and-basic-science/

ebooksecure.com

(eBook PDF) Strategic Healthcare Management: Planning and

Execution, Second Edition

https://ebooksecure.com/product/ebook-pdf-strategic-healthcare-
management-planning-and-execution-second-edition/

ebooksecure.com

Respiratory Physiology (Mosby Physiology Series) 2nd

Edition Edition Michelle M. Cloutier - eBook PDF

https://ebooksecure.com/download/respiratory-physiology-mosby-
physiology-series-ebook-pdf/

ebooksecure.com

(eBook PDF) Financial Accounting 11th Edition by Jerry J.

Weygandt

https://ebooksecure.com/product/ebook-pdf-financial-accounting-11th-
edition-by-jerry-j-weygandt/

ebooksecure.com
(eBook PDF) Writing That Works: Communicating Effectively
on the Job 12fth Edition

https://ebooksecure.com/product/ebook-pdf-writing-that-works-
communicating-effectively-on-the-job-12fth-edition/

ebooksecure.com
 Contents vii

Phase-Lead Compensation 284

Phase-Lead Design Procedure 285
Lag-Lead Compensation 293
Integration and Differentiation Filters 297
Pid Controllers 299
Pid Controller Design 303
Design by Root Locus 311
Summary 324
References and Further Readings 324 • Problems 325

Chapter 9 Pole-Assignment Design and State Estimation 333

Introduction 333
Pole Assignment 333
State Estimation 342
Observer Model 342
Errors in Estimation 344
Error Dynamics 344
Controller Transfer Function 349
Closed-Loop Characteristic Equation 352
Closed-Loop State Equations 353
Reduced-Order Observers 354
Current Observers 359
Controllability and Observability 364
Systems with Inputs 368
Summary 373
References and Further Readings 374 • Problems 374

Chapter 10 System Identification of Discrete-Time Systems 380

Introduction 380
Identification of Static Systems 381
Identification of Dynamic Systems 384
Black-Box Identification 384
Least-Squares System Identification 391
Estimating Transfer Functions with Partly Known Poles and Zeros 397
Recursive Least-Squares System Identification 399
Practical Factors for Identification 402
Choice of Input 402
Choice of Sampling Frequency 403
Choice of Signal Scaling 403
Summary 404
References and Further Readings 404 • Problems 404
viii Contents

Chapter 11 Linear Quadratic Optimal Control 408

Introduction 408
The Quadratic Cost Function 409
The Principle of Optimality 411
Linear Quadratic Optimal Control 414
The Minimum Principle 423
Steady-State Optimal Control 424
Optimal State Estimation—Kalman Filters 430
Least-Squares Minimization 436
Summary 436
References and Further Readings 437 • Problems 438

Chapter 12 Case Studies 444

Introduction 444
Servomotor System 445
System Model 446
Design 449
Environmental Chamber Control System 451
Temperature Control System 453
Aircraft Landing System 457
Plant Model 458
Design 458
Neonatal Fractional Inspired Oxygen 464
Plant Transfer Function 464
Taube’s PID Controller 466
MATLAB pidtool PIDF Controllers 467
Topology Identification in Electric Power System Models 474
References 478

Appendix I Design Equations 480

Appendix II Mason’s Gain Formula 481
Appendix III Evaluation of E*(s) 486
Appendix IV Review of Matrices 491
Appendix V The Laplace Transform 498
Appendix VI z-Transform Tables 512

Index 515
 Preface

This book is intended to be used primarily as a text for a first course in discrete-time control
systems at either the senior undergraduate or first-year graduate level. Furthermore, the text is
suitable for self-study by the practicing control engineer.
This book is based on material taught at both Auburn University and North Carolina State
University, and in intensive short courses taught in both the United States and Europe. The
practicing engineers who attended these short courses have influenced both the content and the
direction of this book, resulting in emphasis placed on the practical aspects of designing and
implementing digital control systems.
Chapter 1 presents a brief introduction and an outline of the text. Chapters 2–11 cover
the analysis and design of discrete-time linear control systems. Some previous knowledge
of ­continuous-time control systems is helpful in understanding this material. The mathemat-
ics involved in the analysis and design of discrete-time control systems is the z-transform and
­vector-matrix difference equations; these topics are presented in Chapter 2. Chapter 3 is devoted
to the very important topic of sampling signals and the mathematical model of the sampler and
data hold. This model is basic to the remainder of the text. The implications and the limitations
of this model are stressed.
The next four chapters, 4–7, are devoted to the application of the mathematics of Chapter 2
to the analysis of discrete-time systems, emphasis on digital control systems. Classical design
techniques are covered in Chapter 8, with the frequency-response Bode technique emphasized.
Modern design techniques are presented in Chapters 9–11. Chapter 12 summarizes some case
studies in discrete-time control system design. Throughout these chapters, practical computer-
aided analysis and design using MATLAB are stressed.
In this fourth edition, several changes have been made. We
• Added additional MATLAB examples throughout the chapters.
• Added a new chapter on system identification (Chapter 11).
• Added new problems in many of the chapters.
• Renumbered the end-of-chapter problems to reflect their corresponding textbook sections.
• Added the MATLAB pidtool design technique in Chapter 8.
• Added two new case studies in Chapter 12.
• Removed four chapters (formerly Chapters 11–14) and two appendices (formerly
Appendices V and VI) on digital filter implementation to reduce the overall page count,
thus placing more emphasis on control design.
Each end-of-chapter problem has been written to illustrate basic material in the chapter.
Generally, short MATLAB programs are given with many of the textbook examples to illustrate
the computer calculations of the results of the example. These programs are easily modified for
the homework problems.
To further assist instructors using this book, a set of PowerPoint slides and a manual con-
taining problem solutions has been developed. The authors feel that the problems at the end of
the chapters are an indispensable part of the text, and should be fully utilized by all who study
this book. Requests for both the problem solutions and PowerPoint slides can be sent directly to
the publisher.

ix
x Preface

At Auburn University, three courses based on the controls portion of this text, Chapters 2–11,
have been taught. Chapters 2–8 are covered in their entirety in a one-quarter four-credit-hour
graduate course. Thus the material is also suitable for a three-semester-hour course and has been
presented as such at North Carolina State University. These chapters have also been covered in
twenty lecture hours of an undergraduate course, but with much of the material omitted. The
topics not covered in this abbreviated presentation are state variables, the modified z-transform,
nonsynchronous sampling, and closed-loop frequency response. A third course, which is a one-
quarter three-credit course, requires one of the above courses as a prerequisite, and introduces the
state variables of Chapter 2. Then the state-variable models of Chapter 4, and the modern design
of Chapters 9–11, are covered in detail. In a recent offering at North Carolina State University,
Chapters 2–11 were covered in a one-semester, three-credit-hour course using this new edition
and the companion set of PowerPoint slides that are also available.
Finally, we gratefully acknowledge the many colleagues, graduate and undergraduate stu-
dents, and staff members of the Electrical Engineering Department at Auburn University who
contributed to the development of the first three editions of this book. In particular, we wish to
thank Professor J. David Irwin, Electrical Engineering Department Head at Auburn University,
for his aid and encouragement during those years. We would also like to acknowledge our col-
leagues and students in the Electrical and Computer Engineering Department at North Carolina
State University for their contributions to and support for this fourth edition.

Charles L. Phillips
Auburn University
H. Troy Nagle
North Carolina State
University
Aranya Chakrabortty
North Carolina State
University
▪ ▪ ▪ ▪ ▪
1
Introduction

1.1 Overview
This book is concerned with the analysis and design of closed-loop physical systems that contain
digital computers. The computers are placed within the system to modify the dynamics of the
closed-loop system such that a more satisfactory system response is obtained.
A closed-loop system is one in which certain system forcing functions (inputs) are deter-
mined, at least in part, by the response (outputs) of the system (i.e., the input is a function of the
output). A simple closed-loop system is illustrated in Fig. 1-1. The physical system (process)
to be controlled is called the plant. Usually a system, called the control actuator, is required to
drive the plant; in Fig. 1-1 the actuator has been included in the plant. The sensor (or sensors)
measures the response of the plant, which is then compared to the desired response. This differ-
ence signal initiates actions that result in the actual response approaching the desired response,
which drives the difference signal toward zero. Generally, an unacceptable closed-loop
response occurs if the plant input is simply the difference between the desired response and
actual response. Instead, this difference signal must be processed (filtered) by another physi-
cal system, which is called a compensator, a controller, or simply a filter. One problem of the
control system designer is the design of the compensator.
An example of a closed-loop system is the case of a pilot landing an aircraft. For this
example, in Fig. 1-1 the plant is the aircraft and the plant inputs are the pilot’s manipulations of
the various control surfaces and of the aircraft velocity. The pilot is the sensor, with his or her
visual perceptions of position, velocity, instrument indications, and so on, and with his or her
sense of balance, motion, and so on. The desired response is the pilot’s concept of the desired
flight path. The compensation is the pilot’s manner of correcting perceived errors in flight path.
Hence, for this example, the compensation, the sensor, and the generation of the desired response
are all functions performed by the pilot. It is obvious from this example that the compensation
must be a function of plant (aircraft) dynamics. A pilot trained only in a fighter aircraft is not
qualified to land a large passenger aircraft, even if he or she can manipulate the controls.
We will consider systems of the type shown in Fig. 1-1, in which the sensor is an appro-
priate measuring instrument and the compensation function is performed by a digital computer.
1
2 Chapter 1 • Introduction

Desired Difference Plant

response signal input Response
Compensation Plant
+
-

Sensor

Figure 1-1 Closed-loop system.

The plant has dynamics; we will program the computer such that it has dynamics of the same
nature as those of the plant. Furthermore, although generally we cannot choose the dynamics
of the plant, we can choose those of the computer such that, in some sense, the dynamics of
the closed-loop system are satisfactory. For example, if we are designing an automatic aircraft
landing system, the landing must be safe, the ride must be acceptable to the pilot and to any pas-
sengers, and the aircraft cannot be unduly stressed.
Both classical and modern control techniques of analysis and design are developed in this
book. Almost all control-system techniques developed are applicable to linear time-invariant
discrete-time system models. A linear system is one for which the principle of superposition
applies [1]. Suppose that the input of a system x1(t) produces a response (output) y1(t), and the
input x2(t) produces the response y2(t). Then, if the system is linear, the principle of superposi-
tion applies and the input [a1x1(t) + a2x2(t)] will produce the output [a1 y1(t) + a2 y2(t)], where
a1 and a2 are any constants. All physical systems are inherently nonlinear; however, in many
systems, if the system signals vary over a narrow range, the system responds in a linear manner.
Even though the analysis and design techniques presented are applicable to linear systems only,
certain nonlinear effects will be discussed.
When the parameters of a system are constant with respect to time, the system is called a
time-invariant system. An example of a time-varying system is the booster stage of a space vehicle,
in which fuel is consumed at a known rate; for this case, the mass of the vehicle decreases with time.
A discrete-time system has signals that can change values only at discrete instants of time.
We will refer to systems in which all signals can change continuously with time as continuous-time,
or analog, systems.
The compensator, or controller, in this book is a digital filter. The filter implements a trans-
fer function. The design of transfer functions for digital controllers is the subject of Chapters 2
through 9 and 11. Once the transfer function is known, algorithms for its realization must be
programmed on a digital computer. In Chapter 10 we introduce system identification methods
to model the plant’s dynamic behavior. In Chapter 12 we present several case studies in digital
controls systems design.
Presented next in this chapter is an example digital control system. Then the equations
describing three typical plants that appear in closed-loop systems are developed.

1.2 Digital Control System

The basic structure of a digital control system will be introduced through the example of an auto-
matic aircraft landing system. The system to be described is similar to the landing system that is
currently operational on U.S. Navy aircraft carriers [2]. Only the simpler aspects of the system
will be described.
 1.2 Digital Control System 3

The automatic aircraft landing system is depicted in Fig. 1-2. The system consists of three
basic parts: the aircraft, the radar unit, and the controlling unit. During the operation of this con-
trol system, the radar unit measures the approximate vertical and lateral positions of the aircraft,
which are then transmitted to the controlling unit. From these measurements, the controlling unit
calculates appropriate pitch and bank commands. These commands are then transmitted to the
aircraft autopilots, which in turn cause the aircraft to respond accordingly.
In Fig. 1-2 the controlling unit is a digital computer. The lateral control system, which
controls the lateral position of the aircraft, and the vertical control system, which controls the
altitude of the aircraft, are independent (decoupled). Thus the bank command input affects
only the lateral position of the aircraft, and the pitch command input affects only the alti-
tude of the aircraft. To simplify the treatment further, only the lateral control system will be
discussed.
A block diagram of the lateral control system is given in Fig. 1-3. The aircraft lateral posi-
tion, y(t), is the lateral distance of the aircraft from the extended centerline of the runway. The
control system attempts to force y(t) to zero. The radar unit measures y(t) every 0.05 s. Thus

processes these sampled values and generates the discrete bank commands h(kT ). The data hold,
y(kT) is the sampled value of y(t), with T = 0.05 s and k = 0, 1, 2, 3, . . . . The digital controller

which is on board the aircraft, clamps the bank command h(t) constant at the last value received
until the next value is received. Then the bank command is held constant at the new value until
the following value is received. Thus the bank command is updated every T = 0.05 s, which is
called the sample period. The aircraft responds to the bank command, which changes the lateral
position y(t).
Two additional inputs are shown in Fig. 1-3. These are unwanted inputs, called distur-
bances, and we would prefer that they not exist. The first, w(t), is the wind input, which certainly
affects the position of the aircraft. The second disturbance input, labeled radar noise, is present
since the radar cannot measure the exact position of the aircraft. This noise is the difference

Aircraft

Radar
Transmitter unit

Bank Lateral
command position
Pitch Controlling Vertical
command unit position

Figure 1-2 Automatic aircraft landing system.

4 Chapter 1 • Introduction

Radar
site

y (t )

Runw ay
Runw ay cent er line

(a)

w (t )

h(t ) Aircraft y (t )
lateral
Bank system Aircraft
command position

Data T
hold

y (kT )
+
+ Radar
noise

Radar

Lateral
digital
controller

Desired
position

(b)
Figure 1-3 Aircraft lateral control system.

between the exact aircraft position and the measured position. Since no sensor is perfect, sensor
noise is always present in a control system.
The design problem for this system is to maintain y(t) at a small level in the presence of the
wind and radar-noise disturbances. In addition, the plane must respond in a manner that both is
acceptable to the pilot and does not unduly stress the structure of the aircraft.

wind input w(t), the bank command input h(t), and the lateral position y(t). These mathemati-
To effect the design, it is necessary to know the mathematical relationships between the

cal relationships are referred to as the mathematical model, or simply the model, of the aircraft.
For example, for the McDonnell-Douglas Corporation F4 aircraft, the model of lateral system is

mand h(t) remains small in amplitude, the nonlinearities are not excited and the system model
a ninth-order ordinary nonlinear differential equation [3]. For the case in which the bank com-
 1.3 The Control Problem 5

described by this ninth-order ordinary nonlinear differential equation may be used for design
purposes.
The task of the control system designer is to specify the processing to be accomplished
in the digital controller. This processing will be a function of the ninth-order aircraft model, the
expected wind input, the radar noise, the sample period T, and the desired response characteris-
tics. Various methods of digital controller design are developed in Chapters 8, 9, and 11.
The development of the ninth-order model of the aircraft is beyond the scope of this book.
In addition, this model is too complex to be used in an example in this book. Hence, to illustrate
the development of models of physical systems, the mathematical models of four simple, but
common, control-system plants will be developed later in this chapter. Two of the systems relate
to the control of position, the third relates to temperature control, and the fourth one describes
control of electrical power in single-machine infinite bus models of power systems. In addition,
Chapter 10 presents procedures for determining the model of a physical system from input–­
output measurements of the system.

1.3 The Control Problem

We may state the control problem as follows. A physical system or process is to be accurately
controlled through closed-loop, or feedback, operation. An output variable (signal), called the
response, is adjusted as required by an error signal. The error signal is a measure of the differ-
ence between the system response, as determined by a sensor, and the desired response.
Generally, a controller, or filter, is required to process the error signal in order that certain
control criteria, or specifications, will be satisfied. The criteria may involve, but not be limited to:
1. Disturbance rejection
2. Steady-state errors
3. Transient response
4. Sensitivity to parameter changes in the plant
Solving the control problem will generally involve:
1. Choosing sensors to measure the required feedback signals
2. Choosing actuators to drive the plant
3. Developing the plant, sensor, and actuator models (equations)
4. Designing the controller based on the developed models and the control criteria
5. Evaluating the design analytically, by simulation, and finally, by testing the physical
system
6. Iterating this procedure until a satisfactory physical-system response results
Because of inaccuracies in the mathematical models, the initial tests on the physical system may
not be satisfactory. The controls engineer must then iterate this design procedure, using all tools
available, to improve the system. Intuition, developed while experimenting with the physical
system, usually plays an important part in the design process.
Fig. 1-4 illustrates the relationship of mathematical analysis and design to physical-­system
design procedures [4]. In this book, all phases shown in the Fig. are discussed, but the emphasis
is necessarily on the conceptual part of the p­ rocedures—​­the application of mathematical con-
cepts to mathematical models. In practical design situations, however, the major difficulties are
in formulating the problem mathematically and in translating the mathematical solution back to
the physical world. Many iterations of the procedures shown in Fig. 1-4 are usually required in
practical situations.
Visit https://testbankfan.com
now to explore a rich
collection of testbank or
solution manual and enjoy
exciting offers!
6 Chapter 1 • Introduction

Mathematical
Problem formulation
model of
system

Physical Conceptual
system aspects
Mathematical
solution of
Solution translation mathematical
problem

Figure 1-4 Mathematical solutions for physical systems.

Depending on the system and the experience of the designer, some of the steps listed ear-
lier may be omitted. In particular, many control systems are implemented by choosing standard
forms of controllers and experimentally determining the parameters of the controller; a specified
step-by-step procedure is applied directly to the physical system, and no mathematical models
are developed. This type of procedure works very well for certain control systems. For other
systems, it does not. For example, a control system for a space vehicle cannot be designed in this
manner; this system must perform satisfactorily the first time it is activated.
In this book mathematical procedures are developed for the analysis and design of control
systems. The techniques developed may or may not be of value in the design of a particular con-
trol system. However, standard controllers are utilized in the developments in this book. Thus
the analytical procedures develop the concepts of control system design and indicate applica-
tions of each of the standard controllers.

1.4 Satellite Model

As the first example of the development of the mathematical model of a physical system, we will

the thruster configuration shown in Fig. 1-5. Suppose that w(t) is the yaw angle of the satellite.
consider the attitude control system of a satellite. Assume that the satellite is spherical and has

In addition to the thrusters shown, thrusters will also control the pitch angle and the roll angle,

systems, whose purpose is to control the angle w(t).

giving complete three-axis control of the satellite. We will consider only the yaw-axis control

For the satellite, the thrusters, when active, apply a torque v(t). The torque of the two
active thrusters shown in Fig. 1-5 tends to reduce w(t). The other two thrusters shown tend to
increase w(t).
Since there is essentially no friction in the environment of a satellite, and assuming the
satellite to be rigid, we can write

d 2 w(t)
J = v(t) (1-1)
dt 2
 1.4 Satellite Model 7

w(t )

Thrusters

Thrusters

Figure 1-5 Satellite.

where J is the satellite’s moment of inertia about the yaw axis. We now derive the transfer func-
tion by taking the Laplace transform of (1-1):

Js2Θ(s) = T(s) = l[v(t)] (1-2)

Initial conditions are ignored when deriving transfer functions. Equation (1-2) can be expressed as

Θ(s) 1
= Gp(s) = 2 (1-3)
T(s) Js

The ratio of the Laplace transforms of the output variable [w(t)] to input variable [v(t)] is called
the plant transfer function, and is denoted here as Gp(s). A brief review of the Laplace transform
is given in Appendix V.
The model of the satellite may be specified by either the second-order differential equation
of (1-1) or the second-order transfer function of (1-3). A third model is the state-variable model,
which we will now develop. Suppose that we define the variables x1(t) and x2(t) as

x1(t) = w(t)
x2(t) = x1(t) = w(t)
(1-4)
# #
(1-5)
#
where x1(t) denotes the derivative of x1(t) with respect to time. Then, from (1-1) and (1-5),

x2(t) = w(t) = v(t)

# $ 1
(1-6)
J

where w(t) is the second derivative of w(t) with respect to time.

$
8 Chapter 1 • Introduction

We can now write (1-5) and (1-6) in vector-matrix form (see Appendix IV):

d + C 1 S v(t)
# 0
x1(t) 0 1 x1(t)
c# d = c dc (1-7)
x2(t) 0 0 x2(t)
J

In this equation, x1(t) and x2(t) are called the state variables. Hence we may specify the model of
the satellite in the form of (1-1), or (1-3), or (1-7). State-variable models of analog systems are
considered in greater detail in Chapter 4.

1.5 Servomotor System Model

In this section the model of a servo system (a positioning system) is derived. An example of this
type of system is an antenna tracking system. In this system, an electric motor is utilized to rotate
a radar antenna that tracks an aircraft automatically. The error signal, which is proportional to
the difference between the pointing direction of the antenna and the line of sight to the aircraft, is
amplified and drives the motor in the appropriate direction so as to reduce this error.
A dc motor system is shown in Fig. 1-6. The motor is armature controlled with a constant
field. The armature resistance and inductance are Ra and L a, respectively. We assume that the
inductance L a can be ignored, which is the case for many servomotors. The motor back emf em(t)
is given by [5]

em(t) = Kbx(t) = Kb
dw(t)
(1-8)
dt

where w(t) is the shaft position, x(t) is the shaft angular velocity, and Kb is a motor-dependent

friction. Letting v(t) be the torque developed by the motor, we write

constant. The total moment of inertia connected to the motor shaft is J, and B is the total viscous

d 2w(t)
v(t) = J
dw(t)
+ B (1-9)
dt 2 dt

if = constant

i
Ra La

e em J B

Figure 1-6 Servomotor system.

 1.5 Servomotor System Model 9

The developed torque for this motor is given by

v(t) = KT i(t) (1-10)

where i(t) is the armature current and KT is a parameter of the motor. The final equation required
is the voltage equation for the armature circuit:

e(t) = i(t)Ra + em(t) (1-11)

These four equations may be solved for the output w(t) as a function of the input e(t). First,
from (1-11) and (1-8),

e(t) - em(t) e(t) Kb dw(t)

i(t) = = - (1-12)
Ra Ra Ra dt

Then, from (1-9), (1-10), and (1-12),

d 2w(t)
v(t) = KT i(t) =
KT KT K b dw(t) dw(t)
e(t) - = J + B (1-13)
Ra Ra dt dt 2 dt

This equation may be written as

d 2w(t) BRa + KT Kb dw(t) KT

J + = e(t) (1-14)
dt 2 Ra dt Ra

which is the desired model. This model is second order; if the armature inductance cannot be
neglected, the model is third order [6].
Next we take the Laplace transform of (1-14) and solve for the transfer function:

Θ(s) KT >Ra KT >JRa

= Gp(s) = = (1-15)
E(s) BRa + K K
T b BRa + KT Kb
Js2 + s s as + b
Ra JRa

Many of the examples of this book are based on this transfer function.
The state-variable model of this system is derived as in the preceding section.
Let

x1(t) = w(t)
x2(t) = w(t) = x1(t)
# # (1-16)

Then, from (1-14),

x2(t) = w(t) = -
# $ BRa + KT Kb KT
x2(t) + e(t) (1-17)
JRa JRa
10 Chapter 1 • Introduction

Hence the state equations may be written as

# 0 1 0
x1(t) x (t)
c# d = C BRa + KT Kb S c 1 d + C KT S e(t) (1-18)
x2(t) 0 - x2(t)
JRa JRa

Antenna Pointing System

We define a servomechanism, or more simply, a servo, as a system in which mechanical posi-
tion is controlled. Two servo systems, which in this case form an antenna pointing system, are

yaw angle, w(t), is controlled by the electric motor and gear system (the control actuator) shown
illustrated in Fig. 1-7. The top view of the pedestal illustrates the yaw-axis control system. The

in the side view of the pedestal.

Side view
of pedestal

Antenna h(t )
Difference Power
amplifier amplifier Gears
vi Motor
vi - vo voltage
Voltage Motor
proportional Error
to desired
angle Shaft
vo encoder
Data
Voltage hold
proportional
to angle Binary

w(t )
Motor code
shaft

Top view of pedestal

(a)

Servo Output
Amplifier motor system
+ Error K1 24
K
s(s + a)
-

-4.8 4.8 Input

Sensor -24

(b) (c)
Figure 1-7 Servo control system.
Exploring the Variety of Random
Documents with Different Content
Armour Car-Lines and on shipments going beyond Chicago the
rebate that seemed necessary to get business was $35 a car. So Mr.
Leeds, the manager of the Santa Fe car-line testified in April 1904
before the Interstate Commerce Commission. Part of Mr. Leed’s
testimony in answer to the questions of the Commission and of its
counsel Mr. Marchand was as follows:[288]
“Mr. Leeds. This is the first year that we entered into the
deciduous fruit business in Northern California, and I met the
competition which we found there when we began business.
“Mr. Marchand. What competition?
“Mr. Leeds. I think it amounts to $25 a car.
“Mr. Marchand. $25 a car?
“Mr. Leeds. Yes, sir.
“Mr. Marchand. By whom?
“Mr. Leeds. We had only one competition.
“Mr. Marchand. Who was your competitor?
“Mr. Leeds. The Armour Car-Line.
“Mr. Marchand. And it was necessary to give $25 or more in
order to secure the traffic—was that your idea?
“Mr. Leeds. I believed so.
“Commissioner Clements. Uniformly $25 a car?
“Mr. Leeds. I think there would be some exception, as to business
farther east than Chicago.
“Commissioner Clements. Would it be more than that?
“Mr. Leeds. Yes, sir.
“Commissioner Clements. What on Eastern business?
“Mr. Leeds. An additional $10.
“Commissioner Clements. $35?
“Mr. Leeds. Yes, sir.
“Commissioner Clements. You pay $25 back to Chicago and
points west of Chicago?
“Mr. Leeds. Yes, sir.
“Commissioner Clements. And $35 to points east of Chicago?
“Mr. Leeds. That is what it would amount to.
 “Commissioner Prouty. Do you agree to do that before the
shipment is made, or afterwards?
“Mr. Leeds. Before.
“Commissioner Prouty. Are your agents authorized to make that
discount?
“Mr. Leeds. No; they are not.
“Commissioner Prouty. Where is the agreement made, and with
whom?
“Mr. Leeds. Myself.
“Commissioner Prouty. Do your agents there know anything
about it?
“Mr. Leeds. I do not think they know what it is. They may know
that something of that kind is going on, but not what it amounts to.
“Commissioner Clements. How does the shipper know that he
can get this $25 and $35 back?
“Mr. Leeds. Well, he probably could not ship if he did not know it.
“Commissioner Clements. How does he find it out? You say your
agents there do not inform him.
“Mr. Leeds. Well, I spent about three months there in the past
year.
“Commissioner Clements. You have advised them all that that
was done, have you?
“Mr. Leeds. We sought the business.”
Mr. Watson appears to have received on California shipments
about $50,000 a year in rebates from the Fruit Growers’ Express
(now an Armour line), and perhaps the amount was nearer
$100,000.[289]
The reduction of icing charges to favored shippers is, of course,
only another way of paying rebates. Yet the car-lines contend that
icing charges are compensation for a private service which is not part
of the transportation service, and therefore outside the Interstate
law. The Interstate Commerce Commission says: “It has been very
customary in the past, and the practice still prevails in some
quarters, to allow to particular shippers a reduction in these
refrigerator charges. Testimony recently taken at Chicago shows that
one large shipper of California to various eastern destinations was
allowed concessions of this kind, which probably aggregated in a
series of seven or eight years several hundred thousand dollars.”[290]
The testimony of H. J. Streychmans before the Commission at
Chicago, May 12, 1905, throws much light on the Armour Car
business. Mr. Streychmans was for over 4 years, from April, 1900, to
August 1904, in the employ of Armour & Company, and the Fruit
Growers’ Express, one of their car-line systems. One of his duties was
to check ice bills. He says the Armour Car-Lines generally pay $2 to
$2.50 a ton for ice, except on the St. Paul and Northwestern and
Erie. On the Northwestern the Armours paid $1 a ton for ice, and on
the Erie $1.25 or $1.50. “These were the main lines. The
Northwestern and St. Paul handled practically all the green fruit
shipments, and the Erie used to get the shipments east.” The profits
were “five or six hundred percent.” On the very long hauls the
percentage was not so high. From Fresno, California, to Boston, for
example, the cost of icing was about $38 and the Armour tariff
charge for icing was $125, leaving a margin of $87 a car.
On some roads Streychmans says that rebates were paid the
Armours on ice. The Chicago, Milwaukee and St. Paul, for example,
billed the ice at $2.50, but in paying the railroad for the ice the
Armours put in a rebate claim for $1 a ton, reducing the net cost to
$1.50. On the Texas and Pacific, the company furnishing the ice
remitted $1 per ton making the net price $2.50. Ice cold rebates were
also paid at Buffalo.
The Armours in their turn made “allowances” to favored shippers.
Streychmans had to make up “allowance statements” “showing the
number of cars shipped by the shippers and giving him a rate of 60
percent of the tariff rate.” A “rebate of $15 to $25 a car” was paid
back. The last statement Mr. Streychmans put in typewriting before
leaving the Armour service in California was for a rebate of 45
percent to Alden Anderson, Lieutenant-Governor of California. The
witness saw on the office file statements of rebates to the Southern
California Fruit Exchange of $10 a car on 1904 shipments of oranges,
etc. A number of shippers in California got rebates amounting to 45
to 50 percent of the icing charges. They paid the actual cost of icing
plus a bonus of $10 to Chicago, $15 to New York, and $20 to Boston.
The cost and bonus together were ordinarily less than half the tariff
charges. For instance, the Armour ice tariff to Boston from Southern
California was $120, the cost $38, and the bonus $20,—$58 total, or
a little less than half the tariff. The full tariff rates were collected and
the difference paid back. Shippers not in on the secret-rebate
arrangement paid the full rates and got no discount.
From Portland, Ore., to Chicago the Armour icing charge was $45,
because the Northern Pacific cars are there to compete; but further
south, at Medford, Ore., where there is only the Southern Pacific, in
league with the Armours, the icing charge to Chicago is $75.
When possible the car-line runs the cars without ice, sometimes
for long distances, but charges the shippers for icing just as if it had
been done.
Some of the railroads pay a bonus for the Armour business, the St.
Paul, the Northwestern, and the Grand Trunk, for example; in other
words, the Armour lines not only charge extortionate rates for icing
and get a mileage on their cars loaded or empty, but in some cases
sell their tonnage to the railroads. In California, however, the witness
believes there is a traffic commission to settle questions of the
division of traffic between the Santa Fe cars and the Armour cars on
the Southern Pacific.
Mr. Streychmans as a confidential clerk was supplied with a secret
code for use in his correspondence. The inside title-page says:
“Transportation Department, General Offices, 205 La Salle Street,
Chicago, Ill. Cipher code No. 100; for exclusive use between
themselves and H. Streychmans. July 1, 1902. Armour Printing
Works, Chicago.”[291]
Some of the cipher words and their meanings are as follows:—
Launching—Can make rebate.
Laundry—Force payment higher rebates.
Laura—Handle rebate matters very carefully.
Laurus—Pay rebates.
Lava—Pay rebates from cash on hand.
Lavello—Rebate must be confidential.
Lavishment—Working for rebate on.
Kinsley—Shade rates a little rather than lose business.
Apples—What allowance is necessary to secure business.
Joculariss—Divide rate.
 Jewelry—Rates being secretly cut by all lines.
Judiciary—Keep your rates below all others.
Junior—Rates must be made which will secure the business.
Junk—If necessary to secure the shipment you can make the rate
to.
Juvenal—Maintain rates unless others cut.
Kadmaster—Manipulate rates so as to.
Kalatna—Meet any rate offered.
Footpath—Interstate Commerce Commission.
Footprint—Avoid service of summons from I. C. C.
Footrot—Meeting of the I. C. C. at —— on —— to consider question
of ——.
Imprint—Martin A. Knapp of New York, Chairman.
Imprinted—Judson C. Clements of Georgia.
Imprinting—James D. Yeomans of Iowa.
Imprison—Charles A. Prouty of Vermont.
Improbitas—Joseph W. Fifer of Illinois.
Improbity—Edward A. Mosely, Secretary.
Armour—Arrange this with the utmost secrecy.
It is evident that the Armour Car-Lines make a business of
arranging secret rebates, evading the law and eluding the Interstate
Commission.
There are some 300 private car-lines in the country owning and
operating about 130,000 private cars. But the law of concentration is
acting on the private cars as well as on the railways, and the private
cars are rapidly consolidating in few hands. Speaking of this
movement in the refrigerator business, the Interstate Commission
says in its Report for 1904, p. 14: “Some years ago there were a
number of these private-car companies which provided refrigerator
cars for the transportation of fruit under refrigeration. Some of these
were the Fruit Growers’ Express, the Kansas City Fruit Express, the
Continental Fruit Express, and the Armour Refrigerator lines. These
companies were all independent of one another originally, and their
cars were used in competition with each other.... At the present day
all the above car companies have been absorbed by the Armour Car-
Lines Company, which has to-day, in our opinion, a practical
monopoly of the movement of fruit in large quantities in most
sections of the country. There is the American Transit Refrigerator
Company, which operates over the Gould lines, and the Santa Fe
Fruit Express, which operates over the Santa Fe System, and there
are numerous refrigerator lines, having a small number of cars and
engaged in a particular service, but we know of no company other
than the Armour Car-Lines which could move the peach crop of
Georgia or the fruits of Michigan. And this company, having acquired
sufficient strength to do so, has adopted the rule that it will not allow
its cars to go on the line of any railroad for the purpose of moving
fruit from points of origin on that railroad, unless it be under what is
known as an exclusive contract.”
By force of the enormous shipments the Armours control they
have compelled railroad after railroad to make the exclusive
contracts they desire, fix rates at their dictation, collect exorbitant
icing charges, give them an excessive mileage allowance, return their
cars empty if they will at high speed instead of detaining them for
loading back, etc. And “if any railroad dares to disobey their orders
when they impose a requirement it will not get any more of their
traffic. The boycott cannot be visited more effectively upon the
railways. That is the secret of the whole situation. They are the
largest shippers, the most arbitrary, the most remorseless that have
ever been known.”[292]
Is it any wonder that Mr. E. M. Ferguson, representing a dozen
associations of fruit and grocery and produce houses, should tell the
Senate Committee that the “situation is tantamount to commercial
slavery”? “It must be plain to all that commercial freedom in any line
of industry has ceased when a gigantic trust like the Armour interests
are permitted, through ownership and operation of private car-lines
to absolutely control the common highways in so far as the use of
such highways may be required in the transportation of that
particular kind of traffic for which their cars are a necessary
instrumentality of carriage, thus enabling the Armour interests (who,
it will be remembered, are also merchants in the commodities
transported in their cars) to completely dominate over all
independent dealers to the extent of fixing rates, conditions, and
terms under which such independent dealers may use the common
highways.”[293]
The fate of a man left to the mercy of the Armours and the mild
influence of the Sermon on the Mount is similar to the fate of a man
without a gun encountering a tiger in the jungles of Africa. Even the
Government seems to be unable to compel justice in this case. The
big guns of the Federal courts have little or no effect on the packers
and the railroads they have benevolently assimilated. They disobey
injunctions as freely as they do the principles of Christianity and the
dictates of conscience, with the excuse perhaps, as to the last, of lack
of acquaintance.
Standard Oil still practically controls the railroads for the most
part so far as the transportation of oil is concerned, manipulating
rates and service so as to favor its own business and hinder or
destroy the business of competitors.
In the recent examination of Standard Oil methods by the State of
Missouri, L. C. Lohman, for 30 years an oil dealer at Jefferson City,
testified that he had been forced to abandon his dealings with
independent oil companies because the Missouri Pacific and
Missouri, Kansas, and Texas roads refused to accept oil for shipment
to him from these companies.
The railroads discriminate against the Texas oil wells by making
the rates on north-bound oil considerably higher than on south-
bound oil. Again the rate to various points from Lima, the centre of
the Ohio and Indiana oil fields, is considerably higher than from
Chicago, the Standard Oil shipping point. For example:

Miles. Rate per hundred.

Lima to Chattanooga 470 43
Chicago to Chattanooga 643 39.5
Lima to Mobile 916 32.5
Chicago to Mobile 926 23
Lima to New Orleans 962 32.5
Chicago to New Orleans 922 23
Lima to Memphis 512 26.5
Chicago to Memphis 526 18
Lima to Cincinnati 132 10
Chicago to Cincinnati 305 11

It costs 3½ cents more per hundred to ship from Lima, 470 miles,
than from Chicago, 643 miles; 9½ cents more from Lima, 916 miles,
to Mobile, than from Chicago, 926 miles, to the same place. The
shorter distance has the higher rate till you get 50 percent off, then
the half distance from Lima has about the same rate as the 100
percent distance from Chicago.
The average rate on 25 staple commodities is about 2 cents higher
per hundred from Cleveland to New Orleans than from Chicago to
New Orleans, while the rate on petroleum is 8 cents higher. This is a
strong discrimination against the Cleveland refineries in favor of the
Chicago shipping point at Whiting. The Standard Oil is the only
shipper of oil from Whiting.[294]
The methods by which the Standard controls New England are still
in full swing. The report of the Industrial Commission tells how the
Standard Oil railroads keep the independent refineries at Cleveland
out of New England through high rates on oil by rail, while the
Standard ships by water, and by making oil second class unless the
shipper has a private siding or tank opposite the rails of the New
Haven and Hartford Railroad, but fifth class if the shipper has such
siding or tank, i. e., if the shipper is the Standard Oil Co.[295] “The
freight rate from Cleveland to Boston,” says the report, “was formerly
22 cents per hundred pounds alike on iron articles, grain, and
petroleum. But since the Interstate Commerce Act the rates have
been changed, so that the rate on grain is 15 cents per hundred
pounds, on iron 20 cents, and on petroleum 24 cents. Again, on
almost every commodity through rates are made from Cleveland and
other western points to points reached by the New York, New Haven
and Hartford Railroad. On petroleum there are no through rates, but
a local rate is added to the Boston rate. Moreover the New York, New
Haven and Hartford prescribes that petroleum and its products shall
be in the second class of freight unless the person to whom it is
shipped has a private siding or tank opposite the rails, in which case
it is fifth class, the rate for fifth class being probably one-half that for
second class. These arrangements are explainable by the fact that the
Standard Oil Company ships oil from its seaboard refineries to
Boston largely by tank steamers, and distributes it from there for a
comparatively short distance at the local rates.”[296]
In the West the Standard has persuaded the railroads to lift the
rates on oil so high as to make competition difficult. The rate from
Pennsylvania points to Chicago was raised from 17½ cents to 19½
cents, and the rate from Chicago to St. Paul went up from 10 cents to
20 cents.[297] The Standard pumps oil to Chicago by pipe, and the
higher the rates by rail the more impossible it is for the independents
to compete. Of course it is against the direct interests of the railway
stockholders to have rates so high as to check the traffic in oil by rail,
but the Standard does not care about that, and it is a small matter
even to the railroad managers compared to incurring the displeasure
of Standard Oil, which has sufficient control in the railway world to
cause any disobedient railroad most serious loss and even make a
railroad war upon it.
Before the Standard found other methods of controlling
transportation and milking the public it used to receive half a million
dollars a month in rebates. But some railroad men who are in a
position to know say that since 1900 the Standard Oil has not asked
for rebates, the reason being that the tariffs are made in such a way
as to give the Trust all the advantage it requires.[298]
The fight now going on in Kansas between the people and the Oil
Combine has forcibly illustrated the methods of the Standard. When
the Kansas oil fields began to show signs of large prosperity the
Standard went into the State, put up refineries and storage tanks,
laid pipe lines, and began to build a through pipe line from Kansas to
its Chicago station at Whiting. By getting the railroads to raise their
rates on oil, compelling producers to agree to sell their oil only to the
Combine, resorting to cut-throat competition to drive them out of
any market they attempted to enter, they practically captured the oil
business of the State and were able to put the price of crude oil down
and squeeze the independents until many of them were ready to sell
out to the Combine at the victor’s own price.
The power of the Trust over the railroads is illustrated by the case
of Mr. I. E. Knapp of Chanute, who went to the field in 1899 and
secured a number of paying wells. He also obtained a market for his
crude oil with the Omaha and Kansas City gas companies,
transporting the oil in tank cars of his own. In the recent
investigation in Kansas it appeared that he had enlarged his business
till he had 20 tank cars in transit. He paid the railroads 10 cents per
hundred lbs. to Omaha and Kansas City, and they counted the weight
at 6.4 lbs. per gallon. With this rate and ¾ of a cent mileage on his
cars he was able to make a good profit, but suddenly in May, 1902,
two weeks after he had signed a year’s contract with the gas
companies, the railroads changed the weight classification to 7.4 lbs.
per gallon, adding thereby $7.50 per car to the freight, while the
freight on the products of crude remained unchanged. That is, the
Standard could still ship gas-oil as a product of crude at the old
weight of 6.4 lbs. a gallon.[299]
Mr. Knapp protested and the railroad agents, admitting that the
classification was arbitrary and not general even on their own roads,
succeeded in getting the order reversed, but only for a short time,
when back it went, and in reply to further protest from the Kansas
agents their superior officers wrote that they were tired of the
correspondence and declined to discuss the matter further. So for 11
months Mr. Knapp had to fulfil his contract with a handicap of $7.50
per car more cost than he had figured on. The result was that in May,
1903, he turned over his crude oil to the Standard which thereafter
supplied the Omaha and Kansas City gas companies, while Knapp’s
20 cars were side-tracked and in the spring of 1905 were still idle at
Chanute.
The weight classification killed Knapp’s business, but a few small
independents lived in spite of it. So another move was made on the
railroad chess-board. Three great railroads tap the Kansas oil fields:
the Santa Fe, the Missouri, Kansas and Texas, and the Missouri
Pacific. In August, 1904, just as the Standard finished its pipe line to
Kansas City, the rates on crude oil and its products were raised by all
the railroads on the field. The rate to Kansas City went up from 10
cents to 17 cents a hundred; and the rate to St. Louis rose from 15
cents to 22 cents. On a carload of fifty-five thousand lbs. the increase
in the freight to Kansas City was $38.50, or $93.50 total, and $121 to
St. Louis. This was prohibitive. In their testimony given in March last
(1905), shippers, even those who were using their own tank cars,
declared that the change in rates compelled them to stop business at
once and shut down their wells.
The advance in freight was not a part of a general readjustment of
rates. It was made alone. And it made oil rates out of all proportion
to other rates. The freight from Chanute to Kansas City was $50 for a
car of wheat, $40 for corn, $66 for machinery, $28 for cattle, and
$30 for a car of fruit, against $93.50 for oil, the least valuable of all,
and formerly carried for $50 or $55 a car.
 The examiner at the recent Kansas investigation presented the
following letter in explanation of the railroads: “The reason the Santa
Fe and the ‘Katy’ railroads raised rates on oil after the pipe line was
completed was because the Standard’s companies arranged with
them to do so, by agreeing to give them a percentage upon every
barrel of oil that was run through their pipe lines on condition the
railroads would increase the freight rate on oil to a prohibitive rate,
so that all the oil would be forced through the pipe line. Now the
railroads have no oil, but get about ten cents per barrel for all oil
going through the pipe lines.”
This is similar to an arrangement that existed for several years
from 1884 on between the Pennsylvania Railroad and the Oil
Combine by which the railroad was to have a fixed sum per barrel on
26 percent of all the oil going eastward from the Pennsylvania oil
fields, whether the oil went by rail or pipe line,[300] in consideration
of which the railroad was to put up the rates on oil.
In the Kansas case there are other reasons more direct and
powerful perhaps than any traffic arrangement. The Standard people
have acquired a large interest in the Santa Fe. One of their strongest
and most unscrupulous men, H. H. Rogers, has taken a place on the
board of directors. John D. Rockefeller and Wm. Rockefeller are
directors of the Missouri, Kansas and Texas, and the Missouri Pacific
is one of the principal lines of the Gould-Rockefeller system. There
are other indications of the grip the Standard has upon the Kansas
railroads. For example, the Colorado Fuel Company that was so
greatly favored by the Santa Fe is largely owned and managed by the
Standard Oil crowd, and the Standard uses the Santa Fe’s right of
way for its pipe lines in Kansas, and for almost the entire distance
from Kansas City to Whiting.
Kansas has risen in revolt against the Oil Trust, and the
Legislature last year (1905) lowered the freight rates on oil and
passed a bill for the establishment of a State refinery to compete with
the Standard and give the oil producers of the State a chance to
escape from the “commercial tyranny” they are now subjected to in
consequence of the fact that there is practically only one buyer in the
market. The State Supreme Court, however, has decided that the
State refinery act is unconstitutional. The independents might,
however, establish a co-operative refinery of their own and do a good
business, if they could get equal freight rates and sufficient support
from public sentiment to withstand the boycott to which the
Standard would be likely to resort. Only the Standard, it is said, can
get rates that encourage the shipment of oil from Kansas wells at
present. And the Standard custom of putting prices very low where
there is competition, keeping prices high in other regions where
there is no competition, making the people in non-competitive
localities pay the cost of killing competition in other places, is
exceedingly effective, as is also its diabolical habit of ruining
merchants who buy independent oil, by establishing competing
houses close to them and underselling them on the whole line of
goods they handle, the Trust’s wide business enabling it to stand
such losses easily, as the total is only an insignificant fraction of the
profits made in regions where no such fight is in progress.
 CHAPTER XXVII.
THE LONG-HAUL ANOMALY.

The long and short haul clause is still broken by the railroads as
well as by the Supreme Court, especially in the West and in the
South, where the basing-point system causes such grievous
discriminations. For example, with a rate of 48 cents from New York
to Atlanta and a local rate of 38 cents from Atlanta to Suwanee, the
rate from New York to Suwanee is 86 cents, although Suwanee is 31
miles nearer New York than Atlanta. This system is not confined to
places that have water competition. A considerable number of towns
on the Southern Railway and on the Louisville and Nashville have
been made basing-points, though they have no water competition.
[301]

Jacksonville is the main basing-point in Florida, and rates to other

destinations are the rate of Jacksonville plus the local rate from
Jacksonville to destination, even though the destination is nearer the
point of shipment than Jacksonville.[302]
From New Orleans to the “Virginia Cities,” Richmond, Lynchburg,
and Norfolk, is about 800 miles. Charlotte, at the southern border of
North Carolina, is about half way. Yet the rates to Charlotte on a
number of articles are double the rates to the Virginia cities, twice
the distance. The Southern Railway and the Seaboard Air Line reach
the city, but there is no competition. Water competition must be met
in Virginia, but if the Virginia ton-mile rate will pay a profit, is not
the fourfold ton-mile rate to Charlotte an exorbitant charge?[303]
Danville is an excellent example of the evils of place
discrimination. Prior to 1886 Danville enjoyed equal freight rates
with Lynchburg and Richmond through the competition of the
Virginia Midland Railroad and the Southern Railway, but in that year
the Southern road (then known as the Richmond and Danville)
bought the Virginia Midland and deprived Danville of its equal rates.
In 1890 Danville subscribed $100,000 towards the construction of
another competing road, which was built, but after a few years it too
was purchased by the Southern Railway, and the rates were made
strongly adverse to Danville. The matter went to the Commission
and the courts, but the city has not been able to carry on the
litigation with the roads.[304]
The Southern Railway carried bananas in 1902–1903 from
Charleston to Lynchburg for 20 cents a hundred lbs., but if the fruit
stopped at Danville, part way on the road to Lynchburg, the rate was
43 cents a hundred. The road said it had to make low rates at
Lynchburg to meet competing bananas coming in by way of
Baltimore. The Commission found, however, that the Lynchburg rate
was 13 cents lower than the rate justified by competition from
Baltimore or elsewhere.[305] It is claimed that railroad discrimination
has decreased the taxable values of Danville several hundred
thousand dollars from 1900 to 1904. The Danville representative
said, “I have heard a great deal about confiscatory rates, fixed by a
Commission authorized to fix rates, but I have not heard anything
about confiscatory rates fixed by the railroads, whereby the property
of the public and of municipalities and taxable values are destroyed;
but those facts exist. They exist in my town, and these facts exist in
spite of the fact that the city of Danville contributed $100,000 to the
building of the Lynchburg and Danville Railroad.”[306]
The rate on canned goods from Hoopeston, Ill., to Nashville,
Tenn., is 27 cents per hundred. From Hoopeston to Memphis,
several hundred miles further, the rate is 19 cents. From Greenwood,
Ind., to Nashville the rate is 25 cents, to New Orleans 21 cents, to
Mobile 20 cents, and to Memphis 19 cents.[307] The Chesapeake and
Ohio Railway, the Norfolk and Western, and the Baltimore and Ohio,
all carry lumber from the Blue Ridge Mountains. The rate from the
Shenandoah Valley to Philadelphia is 16 cents per hundred, while
from points in the region a hundred miles or so further west the rate
is only 14 cents. “The man who is producing lumber to-day on the
eastern slope of the Blue Ridge Mountains, almost within sight of us,
must pay 2 cents per hundred lbs. more to get lumber to
Philadelphia than the man 50 or 75 miles further west, who gets his
lumber transported for 14 cents. Now, 2 cents a hundred lbs. is 40
cents a ton. That is $12 a carload of 30,000 lbs., and that is probably
about all the margin of profit there is in lumber of that kind.” All
three of the railroads are controlled by one great railroad system, yet
they claim that competition among them justifies the lower rate in
the region where they cross.[308]
The rate on lumber from Chattanooga to Buffalo via Cincinnati is
20 cents, while from Chattanooga to Cleveland, a shorter haul over
the same road, it is 23 cents.[309]
Corn rates now (1905) are 13 cents per hundred from Omaha,
1400 miles to New York, and 25 cents from Boone, Iowa, 1252 miles
to New York, 25 cents also from Dennison, Iowa, 1341 miles to New
York, etc. Many similar facts might be named. And such
discriminations between contiguous markets do not violate the
Interstate Law. There is no requirement that one railroad line shall
not charge less for a given distance than another railroad line
charges, and even on the same line the long and short haul clause
yields to the necessity of meeting competition.
When Dubuque wants to buy things from the South it must pay
much higher rates than Milwaukee, Madison, Chicago, Freeport, etc.
Manufacturers in Fort Dodge and Dubuque, Iowa, have to pay higher
rates to the Pacific than manufacturers in Chicago and the East.
Iowa raises corn, cattle, and hogs, and would like to have packing-
houses, but cannot because of the discrimination in favor of Chicago
and Missouri River points.[310] Iowa business men also say that small
poultry and dressed-meat concerns cannot compete with the big
packers, on account of the private-car system and the concessions
granted the car-lines, and they complain vigorously of the
discrimination against them in the rates on shoes, grain, cattle, iron,
steel, etc. The railroads have decreed that Iowa shall not be a
manufacturing State.
“The Chairman. Why do you say that the railroads have decreed
that Iowa shall not become a manufacturing State?
“Hon. A. B. Cummins, Governor of Iowa. I reach that conclusion
simply because all our manufacturers, when they attempt to reach
beyond our own State, meet rates that so discriminate against them
that they cannot compete with manufacturers elsewhere.”
 In many cases a shipper at an intermediate point between
Minneapolis and Chicago can send his grain to Minneapolis, rebill it
to Chicago, and have it go back through his own town to destination
more cheaply than he can ship direct to destination.[311]
From Cannon Falls the rate to Chicago is 15 cents a hundred on
grain. The rate from Cannon Falls to Minneapolis is 7 cents, and
from Minneapolis to Chicago 7½ cents. So it costs ½ cent a hundred
more to ship from Cannon Falls direct to Chicago than to ship to
Minneapolis and from there back through Cannon Falls to Chicago.
And if he wants to send his grain to Louisville, Ky., it will cost him 5
cents a hundred more to ship from Cannon Falls to Louisville, than if
he sends his grain to Minneapolis and bills it from there to
Louisville.[312]
Denver still suffers from the sort of discrimination described in the
preceding section.[313] The rate in cotton goods from New England to
Denver is $2.24 per hundred. From New England to San Francisco,
1500 miles further on, the rate is $1 a hundred in carload lots. On a
shipment in relation to which a Denver merchant made complaint,
the Burlington road received $25.95 from Chicago to Denver,
whereas if the same shipment had been intended for Frisco the
Burlington would have received only $4.50.
Salt Lake City also is wrestling with adverse freight rates. On
cotton goods the rate from New York to Frisco is $1, while on the
shorter haul from New York to Salt Lake it is just double, $2 per
hundred.[314] The rate on window-shade cloth from New York to Salt
Lake City is $2.30. Carrying it 800 miles further, New York to
California, the railroads charge only $1, and this affords a slight
profit. Is it not clear that the $2.30 is excessive?[315] “The men who
build a city in the interior cannot expect to get as reasonable a rate as
the men who build their city on the shore of the sea, but the
difference should be a reasonable one.”
It would seem that the men who build in the interior might expect
that they would not be called on to pay railway fixed charges on coast
traffic as well as on their own. It is unfair to give the coast people the
celerity of railway traffic at the cost of water traffic. The railroad
theory that every pound of freight is to be secured that will pay the
cost of hauling or a little more, though a water route or a shorter rail
line might carry the freight at less absolute cost, is not in accord with
sound public policy or the saving of industrial power. It is an
economic absurdity to haul by rail what can go more cheaply and as
safely by water. A co-operative company or a consolidated company
of any honest and sensible variety, owning both the railroads and the
steamboat lines, would divide the traffic in such a way as to secure
the maximum economy and convenience, and would make a
reasonable payment for the extra speed and other advantages of
railway transit the main condition of selecting that method of
transportation, with an option in the company under specified
conditions to facilitate the full loading of trains and boats through
the adjustment of rates.
The case of Spokane is a specially aggravated one. The rate on bar
iron from Chicago to Spokane is $2.07 a hundred against $1.25 to
Seattle; iron pipe $1 to Spokane, 50 cents to Seattle; lamps $2.35 to
Spokane, $1.10 to Seattle; belting $3.13 to Spokane, and $1.65 to
Seattle; mining-car wheels $1.26 to Spokane and 85 cents to Seattle;
cottons $1.75 to Spokane, 90 cents to the coast; soap (toilet) $1.23 to
Spokane, 75 cents to coast cities; wire and wire goods $2.35 to
Spokane, $1.50 to the coast; sewing machines $2.25 to Spokane,
$1.40 to coast; typewriters $5.96 to Spokane, $3 to the cities of the
coast.
In general the rates from the East to Spokane are the through rates
to the coast plus the local rates from the coast back to Spokane.[316]
The preference which Tacoma, Seattle, etc., have over Spokane is
about 80 percent. Spokane pays about $1.80 on shipments from
Chicago, while Tacoma and Seattle pay $1.[317] Spokane is a great
railroad junction, but competition has been suppressed by
agreement between the lines, while competition is still active at
Tacoma and Seattle, so that under the decision of the Supreme Court
the railroads are free to discriminate against Spokane. Aside from
water competition the railroads want to build up Seattle. They have
invested a great deal of money in docks and facilities for doing
business there. The manufacture of wooden pipe was flourishing in
Spokane. The company was shipping 2 carloads daily and its pay roll
was $3,000 a month. A rival factory in Seattle, backed by the big
lumber firms of the coast, got the railways to make rates that enabled
it to lay down the manufactured pipe in Spokane about 60 percent
cheaper than the Spokane factory could make it. The situation came
to light November, 1903, two months after the rates went into effect,
when the Spokane factory came into competition with the Seattle
factory for a contract at Butte. The bid of the Seattle firm was less
than the pipe could be sold for at Spokane by the factory in that city,
and Butte is 384 miles east of Spokane. The rates shut off the
Spokane factory from the East entirely. In about 8 months that
flourishing manufacture in Spokane was wiped out.[318] There was no
water competition here to make an excuse for discrimination, for the
cut was made from Seattle east to Spokane and points still further
east.
The paper-box manufacture was forced out of existence in
Spokane by similar discriminations. Eastern factories can lay down
the boxes in Spokane cheaper than the local factories can get the
strawboard. So with other trades. The manufacture of sash would be
rapidly developed if it were not for the grievous discrimination on
window glass, $1.38 from Pittsburg to Spokane, against 90 cents to
Portland, Seattle, etc.
 CHAPTER XXVIII.
OTHER PLACE DISCRIMINATIONS.

There are multitudes of other place discriminations besides those

related to the long and short haul question. The Business Men’s
League of St. Louis and the St. Louis Merchants Exchange complain
of serious discrimination against their city as compared with
Chicago, Kansas City, Omaha, etc., in rates on corn, wheat, oats,
groceries, hardware, and cotton.[319] Des Moines gets supplies from
Chicago at 60 cents, while Fort Dodge, the same distance from
Chicago, pays 72 cents.[320] Shoe manufacturers and wholesale
grocers of Atlanta who have had to close down declare they were
ruined by discriminative freight rates. Two years ago a prohibitive
rate was put on cotton bound for Atlanta, but the freight agents of
the leading railroads entering the city were indicted by the Federal
grand jury and the rate was withdrawn. Mobile complains of loss of
business because of discriminations in favor of New Orleans on one
side and Pensacola on the other. The Fort Wayne Commercial Club
complains of discrimination in rates, demurrage, switching, supply
of cars, etc. The lumber rate to Boston from points in West Virginia
on the Norfolk and Western is 29½ cents, while points on the B. & O.
and Chesapeake and Ohio in the same State and the same distance
from Boston have a rate of 23½ cents.[321] The Pennsylvania Railroad
taking lumber to points on the Long Branch Railroad made the rates
by adding to the New York rate an arbitrary charge of 5 cents a
hundred lbs. if the lumber came from Saginaw, Mich., but only 2
cents if the shipping point was Buffalo; held an unlawful
discrimination.[322]
Even so important a city as Philadelphia has had serious
complaints to make at times of the favoritism shown New York by
sending many of the best trains from Washington north through
Philadelphia without running into Broad Street Station, but stopping
only at West Philadelphia, and by arranging excursion tickets so that
southern buyers would go to New York instead of Philadelphia.[323]
In June, 1905, the New Haven and Hartford notified connecting
lines that it would not receive any further shipments of coal for
delivery east of the Connecticut River or north of Hartford after
August 31. Such an order constitutes a compound discrimination
against certain localities and a specific commodity.
The whole of New England suffers from a discrimination of about
100 percent in freight rates, the average rate in New England being
about double the average for the United States. Quoting my
testimony before the United States Industrial Commission: “Another
phase of discrimination was brought out very prominently in our
studies in New England, and the best source of information, perhaps,
is the report made by the Massachusetts Railroad Commission a few
years ago (1894), in which they compared the average freight rate on
New England roads, individual roads, and the average of all the
roads there, showing that our rates were about double the average
freight rate in the Middle States, or in the Middle West, and that it
was clearly double what the average freight rate was for the whole
United States, and they argued with much force that it was really a
discrimination against New England as a whole, especially against
Boston. One of the pleas put forward in discussing the question of
leasing the Boston and Albany was that the giving over of the Boston
and Albany to the New York Central control would intensify instead
of relieve that sectional discrimination against New England as a
whole, because the road would come under the control of those
interested chiefly in the development of New York City, and not in
the development of Boston and the New England States.”[324]
In the Cincinnati Maximum Rate Case, involving a large number of
railways and steamship lines, the Commission found discrimination
between the rates from the eastern seaboard and central territory to
southern points, and fixed a schedule of maximum rates from
Cincinnati and Chicago to Knoxville, Chattanooga, Rome, Atlanta,
Meridian, Birmingham, Anniston, and Selma, and required the
railroads to revise their rates to other points in the South in
conformity with the provisions of the order.[325] On appeal to the
Supreme Court it was held that the order could not be enforced