Practical methods for optimal control using

nonlinear programming 1st Edition John T. Betts

pdf download

https://ebookname.com/product/practical-methods-for-optimal-
control-using-nonlinear-programming-1st-edition-john-t-betts/

Get Instant Ebook Downloads – Browse at https://ebookname.com

Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

Practical Methods for Optimal Control and Estimation

Using Nonlinear Programming Second Edition Advances in
Design and Control John T. Betts

https://ebookname.com/product/practical-methods-for-optimal-
control-and-estimation-using-nonlinear-programming-second-
edition-advances-in-design-and-control-john-t-betts/

Nonlinear optimal control theory 1st Edition Leonard

David Berkovitz

https://ebookname.com/product/nonlinear-optimal-control-
theory-1st-edition-leonard-david-berkovitz/

Nonparametric Statistical Methods Using R 1st Edition

John Kloke

https://ebookname.com/product/nonparametric-statistical-methods-
using-r-1st-edition-john-kloke/

Beginning Teaching Beginning Learning in Primary

Education 3rd Edition Janet Moyles

https://ebookname.com/product/beginning-teaching-beginning-
learning-in-primary-education-3rd-edition-janet-moyles/
Essential College Physics with MasteringPhysics 1st
Edition Andrew Rex

https://ebookname.com/product/essential-college-physics-with-
masteringphysics-1st-edition-andrew-rex/

Half a Century of Music in England 1837 1887 Essays

Towards a History Francis Hueffer

https://ebookname.com/product/half-a-century-of-music-in-
england-1837-1887-essays-towards-a-history-francis-hueffer/

Examining Core Elements of International Research

Collaboration Institute Of Medicine

https://ebookname.com/product/examining-core-elements-of-
international-research-collaboration-institute-of-medicine/

Methods for the Economic Evaluation of Health Care

Programmes 4th Edition Michael F. Drummond

https://ebookname.com/product/methods-for-the-economic-
evaluation-of-health-care-programmes-4th-edition-michael-f-
drummond/

Fundamentals of Oral Histology and Physiology 1st

Edition Arthur R. Hand

https://ebookname.com/product/fundamentals-of-oral-histology-and-
physiology-1st-edition-arthur-r-hand/
Bruce Lee A Life 1st Edition Matthew Polly

https://ebookname.com/product/bruce-lee-a-life-1st-edition-
matthew-polly/
Practical Methods
for Optimal Control
Using Nonlinear
Programming
Advances in Design and Control

SIAM's Advances in Design and Control series consists of texts and monographs dealing with
all areas of design and control and their applications. Topics of interest include shape
optimization, multidisciplinary design, trajectory optimization, feedback, and optimal control.
The series focuses on the mathematical and computational aspects of engineering design and
control that are usable in a wide variety of scientific and engineering disciplines.

Editor-in-Chief
John A. Burns, Virginia Polytechnic Institute and State University

Editorial Board
H. Thomas Banks, North Carolina State University
Stephen L. Campbell, North Carolina State University
Eugene M. Cliff, Virginia Polytechnic Institute and State University
Ruth Curtain, University of Groningen
Michel C. Delfour, University of Montreal
John Doyle, California Institute of Technology
Max D. Gunzburger, Iowa State University
Rafael Haftka, University of Florida
Jaroslav Haslinger, Charles University
J. William Helton, University of California at San Diego
Art Krener, University of California at Davis
Alan Laub, University of California at Davis
Steven I. Marcus, University of Maryland
Harris McClamroch, University of Michigan
Richard Murray, California Institute of Technology
Anthony Patera, Massachusetts Institute of Technology
H. Mete Soner, Carnegie Mellon University
Jason Speyer, University of California at Los Angeles
Hector Sussmann, Rutgers University
Allen Tannenbaum, University of Minnesota
Virginia Torczon, William and Mary University

Series Volumes
Betts, John T., Practical Methods for Optimal Control Using Nonlinear Programming
El Ghaoui, Laurent and Niculescu, Silviu-lulian, eds., Advances in Linear Matrix Inequality
Methods in Control
Helton, J. William and James, Matthew R., Extending H°° Control to Nonlinear Systems:
Control of Nonlinear Systems to Achieve Performance Objectives
Practical Methods
for Optimal Control
Using Nonlinear
Programming

John T. Betts
The Boeing Company
Seattle, Washington

Siam
Society for Industrial and Applied Mathematics
Philadelphia
Copyright ©2001 by the Society for Industrial and Applied Mathematics.

10987654321

All rights reserved. Printed in the United States of America. No part of this book may
be reproduced, stored, or transmitted in any manner without the written permission
of the publisher. For information, write to the Society for Industrial and Applied
Mathematics, 3600 University City Science Center, Philadelphia, PA 19104-2688.

Library of Congress Cataloging-in-Publication Data

Betts, John T. 1943-

Practical methods for optimal control using nonlinear programming / John T. Betts.
p. cm.— (Advances in design and control)
Includes bibliographical references and index.
ISBN 0-89871-488-5
1. Control theory. 2. Mathematical optimization. 3. Nonlinear programming. I. Series.

QA402.3 .B47 2001

629.8'312-dc21
00-069809

Siam is a registered trademark.

For Theon and Dorothy
This page intentionally left blank
Contents

Preface ix

1 Introduction to Nonlinear Programming 1

1.1 Preliminaries 1
1.2 Newton's Method in One Variable 2
1.3 Secant Method in One Variable 3
1.4 Newton's Method for Minimization in One Variable 5
1.5 Newton's Method in Several Variables 7
1.6 Unconstrained Optimization 8
1.7 Recursive Updates 10
1.8 Equality-Constrained Optimization 12
1.9 Inequality-Constrained Optimization 15
1.10 Quadratic Programming 17
1.11 Globalization Strategies 20
1.12 Nonlinear Programming 27
1.13 An SQP Algorithm 28
1.14 What Can Go Wrong 30

2 Large, Sparse Nonlinear Programming 37

2.1 Overview: Large, Sparse NLP Issues 37
2.2 Sparse Finite Differences 38
2.3 Sparse QP Subproblem 40
2.4 Merit Function 42
2.5 Hessian Approximation 44
2.6 Sparse SQP Algorithm 45
2.7 Defective Subproblems 48
2.8 Feasible Point Strategy 49
2.9 Computational Experience 52
2.10 Nonlinear Least Squares 56

3 Optimal Control Preliminaries 61

3.1 The Transcription Method 61
3.2 Dynamic Systems 61
3.3 Shooting Method 62
3.4 Multiple Shooting Method 63
3.5 Initial Value Problems 65
3.6 Boundary Value Example 72

vii
viii CONTENTS

3.7 Dynamic Modeling Hierarchy 75

3.8 Function Generator 76

4 The Optimal Control Problem 81

4.1 Introduction 81
4.2 Necessary Conditions for the Discrete Problem 84
4.3 Direct versus Indirect Methods 85
4.4 General Formulation 87
4.5 Direct Transcription Formulation 89
4.6 NLP Considerations—Sparsity 92
4.7 Mesh Refinement 107
4.8 Scaling 121
4.9 Quadrature Equations 123
4.10 What Can Go Wrong 125

5 Optimal Control Examples 133

5.1 Space Shuttle Reentry Trajectory 133
5.2 Minimum Time to Climb 138
5.3 Low-Thrust Orbit Transfer 147
5.4 Two-Burn Orbit Transfer 152
5.5 Industrial Robot 165
5.6 Multibody Mechanism 170

Appendix: Software 177

A.I Simplified Usage Dense NLP 177
A.2 Sparse NLP with Sparse Finite Differences 177
A.3 Optimal Control Using Sparse NLP 178

Bibliography 181
Index 189
Preface

Solving an optimal control problem is not easy. Pieces of the puzzle are found scattered
throughout many different disciplines. Furthermore, the focus of this book is on practical
methods, that is, methods that I have found actually work! In fact everything described
in this book has been implemented in production software and used to solve real optimal
control problems. Although the reader should be proficient in advanced mathematics, no
theorems are presented.
Traditionally, there are two major parts of a successful optimal control solution tech-
nique. The first part is the "optimization" method. The second part is the ''differential
equation'' method. When faced with an optimal control problem, it is tempting to sim-
ply "paste" together packages for optimization and numerical integration. While naive
approaches such as this may be moderately successful, the goal of this book is to suggest
that there is a better way! The methods used to solve the differential equations and
optimize the functions are intimately related.
The first two chapters of this book focus on the optimization part of the problem.
In Chapter 1, the important concepts of nonlinear programming for small, dense appli-
cations are introduced. Chapter 2 extends the presentation to problems that are both
large and sparse. Chapters 3 and 4 address the differential equation part of the prob-
lem. Chapter 3 introduces relevant material in the numerical solution of differential (and
differential-algebraic) equations. Methods for solving the optimal control problem are
treated in some detail in Chapter 4. Throughout the book, the interaction between op-
timization and integration is emphasized. Chapter 5 presents a collection of examples
that illustrate the various concepts and techniques.
The book does not cover everything. Many important topics are simply not discussed
in order to keep the overall presentation concise and focused. The discussion is general
and presents a unified approach to solving optimal control problems. Most of the ex-
amples are drawn from my experience in the aerospace industry. Examples have been
solved using a particular implementation called SOCS. I have tried to adhere to nota-
tional conventions from both optimization and control theory whenever possible. Also,
I have attempted to use consistent notation throughout the book.
The material presented here represents the collective contributions of many people.
The nonlinear programming material draws heavily on the work of John Dennis, Roger
Fletcher, Philip Gill, Walter Murray, Michael Saunders, and Margaret Wright. The ma-
terial on differential-algebraic equations is drawn from the work of Uri Ascher. Kathy
Brenari, Steve Campbell, and Linda Petzold. I was introduced to optimal control by
Stephen Citron, and routinely refer to the text by Bryson and Ho [35]. Over the past 10
years I have been fortunate to participate in workshops at Oberwolfach. Munich, Min-
neapolis. Victoria, Lausanne, and Griefswald. I benefited immensely simply by talking
with Larry Biegler, Hans Georg Bock, Roland Bulirsch, Rainer Callies, Kurt Chudej,

ix
x PREFACE

Tim Kelley, Bernd Kugelmann, Helmut Maurer, Rainer Mehlhorn, Angelo Miele, Hans
Josef Pesch, Ekkehard Sachs, Gottfried Sachs, Roger Sargent. Volker Schulz, Mark Stein-
bach, Oskar von Stryk, and Klaus Well.
Two coworkers deserve special thanks. Paul Frank has played a major role in the im-
plementation and testing of the large, sparse nonlinear programming methods described.
Bill Huffman, my coauthor for many publications and the SOCS software, has been an
invaluable sounding board over the last two decades. Finally, thanks to Jennifer for her
patience and understanding during the preparation of this book.

John T. Belts
Chapter 1

Introduction to Nonlinear
Programming

1.1 Preliminaries
This book concentrates on numerical methods for solving the optimal control problem.
The fundamental principle of all effective numerical optimization methods is to solve
a difficult problem by solving a sequence of simpler subproblems. In particular, the
solution of an optimal control problem will require the solution of one or more finite-
dimensional subproblems. As a prelude to our discussions on optimal control, this chapter
will focus on the nonlinear programming (NLP) problem. The NLP problem requires
finding a finite number of variables such that an objective function or performance index
is optimized without violating a set of constraints. The NLP problem is often referred to
as parameter optimization. Important special cases of the NLP problem include linear
programming (LP). quadratic programming (QP), and least squares problems.
Before proceeding further, it is worthwhile to establish the notational conventions
used throughout the book. This is especially important since the subject matter covers
a number of different disciplines, each with their own notational conventions. Our goal
is to present a unified treatment of all these fields. As a rule, scalar quantities will be
denoted by lowercase letters (e.g., a). Vectors will be denoted by boldface lowercase
letters and will usuallv be considered column vectors, as in

where the individual components of the vector are Xk for fc = 1 , . . . , n. To save space, it
will often be convenient to define the transpose, as in

A sequence of vectors will often be denoted as Xfe, x ^ . + i , . . . . Matrices will be denoted by

1
2 CHAPTER 1. INTRODUCTION TO NONLINEAR PROGRAMMING

boldface capital letters, as in

1.2 Newton's Method in One Variable

The fundamental approach to most iterative schemes was suggested over 300 years ago by
Newton. In fact, Newton's method is the basis for all of the algorithms we will describe.
We begin with the simplest form of Newton's method and then in subsequent sections
generalize the discussion until we have presented one of the most widely used NLP
algorithms, namely the sequential quadratic, programming (SQP) method.
Suppose it is required to find the value of the variable x such that the constraint
function

Let us denote the solution by x* and let us assume x is a guess for the solution. The
basic idea of Newton's method is to approximate the nonlinear function c(x) by the first
two terms in a Taylor series expansion about the current point x. This yields a linear
approximation for the constraint function at the new point x, which is given by

where c'(x) = dc/dx is the slope of the constraint at x. Using this linear approximation,
it is reasonable to compute x, a new estimate for the root, by solving (1.5) such that
c(x) = 0, i.e.,

Typically, we denote p = x — x and rewrite (1.6) as

where

Of course, in general, c(x) is not a linear function of x. and consequently we cannot

expect that c(x) = 0. However, we might hope that x is a better estimate for the root
x* than the original guess x, in other words we might expect that

and also

If the new point is an improvement, then it makes sense to repeat the process, thereby
defining a sequence of points x^ ,x^l\x^\... with point (k+l) in the sequence given by
1.3. SECANT METHOD IN ONE VARIABLE 3

For notatioiial convenience, it usually suffices to present a single step of the algorithm, as
in (1.6), instead of explicitly labeling the information at step k using the superscript, no-
tation x^. Nevertheless, it should be understood that the algorithm defines a sequence
of points x p ^°', x' 1 ', x ^ 2 ' , . . .. The sequence is said to converge to x* if

In practice, of course, we are not interested in letting k —> oo. Instead we are satisfied
with terminating the sequence when the computed solution is "close'' to the answer.
Furthermore, the rate of convergence is of paramount importance when measuring the
computational efficiency of an algorithm. For Newton's method, the rate of convergence
is said to be quadratic or, more precisely, q-quadratic (cf. [46]). The impact of quadratic
convergence can be dramatic. Loosely speaking, it implies that each successive estimate
of the solution will double the number of significant digits!
Example 1.1. To demonstrate, let us suppose we want to solve the constraint

where the coefficients 01,02,03 are chosen such that c(O.l) = —0.05, c(0.25) ~ 0, and
c(0.9) = 0.9. Table 1.1 presents the Newton iteration sequence beginning from the initial
guess x = 0.85 and proceeding to the solution at x* = 0.25. Figure 1.1 illustrates the first
three iterations. Notice in Table 1.1 that the error between the computed solution and
the true value, which is tabulated in the third column, exhibits the expected doubling
in significant figures from the fourth iteration to convergence.
So what is wrong with Newton's method? Clearly, quadratic convergence is a very-
desirable property for an algorithm to possess. Unfortunately, if the initial guess is not
sufficiently close to the solution, i.e., within the region of convergence. Newton's method
may diverge. As a simple example, Dennis and Schnabel [46] suggest applying Newton's
method to solve c(x) — arctan(x) = 0. This will diverge when the initial guess x^\ > a,
converge when x^0> < a, and cycle indefinitely if \x^\ — a, where a — 1.3917452002707.
In essence, Newton's method behaves well near the solution (locally) but lacks something
permitting it to converge globally. So-called globalization techniques, aimed at- correcting
this deficiency, will be discussed in subsequent sections. A second difficulty occurs when
the slope c!(x) = 0. Clearly, the correction defined by (1.6) is not well defined in this
case. In fact. Newton's method loses its quadratic convergence property if the slope is
zero at the solution, i.e., c'(x*) = 0. Finally. Newton's method requires that the slope
c'(x) can be computed at every iteration. This may be difficult and/or costly especially
when the function c(x) is complicated.

1.3 Secant Method in One Variable

Motivated by a desire to eliminate the explicit calculation of the slope, one can consider
approximating it at xk by the secant

Notice that this approximation is constructed using two previous iterations, but only
requires values for the constraint function c(x). This expression can be rewritten to give
4 CHAPTER 1. INTRODUCTION TO NONLINEAR PROGRAMMING

Iter. c(x) X IT — Jj
|X T*|
|

1 0.79134615384615 0.85000000000000 0.60000000000000

2 0.18530192382759 0.47448669201521 0.22448669201521
3 3.5942428588261 xlO~2 0.30910437279376 5.9104372793756 x!0~2
4 3.6096528286200xlO~3 0.25669389900972 6.6938990097217xlO-3
5 5. 7007630268141 xlCT5 0.25010744198003 1.0744198002549xlO~4
6 1.5161639596584xlO~8 0.25000002858267 2.8582665845267xlO~8
7 1.05471 18733939 x 10~15 0.25000000000000 1 .8873791418628 x 10~15

Table 1.1: Newton's method for root finding.

Figure 1.1: Newton's method for root finding.

the so-called secant condition

where B is the (scalar) secant approximation to the slope. Using this approximation, it
then follows that the Newton iteration (1.6) is replaced by the secant iteration

which is often written as

Figure 1.2 illustrates a secant iteration applied to Example 1.1 described in the pre-
vious section.
1.4. NEWTON'S METHOD FOR MINIMIZATION IN ONE VARIABLE 5

Figure 1.2: Secant method for root finding.

Clearly, the virtue of the secant method is that it does not require calculation of the
slope c'(xk). While this may be advantageous when derivatives are difficult to compute,
there is a downside! The secant method is superlinearly convergent, which, in general,
is not as fast as the quadratically convergent Newton algorithm. Thus, we can expect
convergence will require more iterations, even though the cost per iteration is less. A
distinguishing feature of the secant method is that the slope is approximated using
information from previous iterates in lieu of a direct evaluation. This is the simplest ex-
ample of a so-called quasi-Newton method.

1.4 Newton's Method for Minimization in One Vari-

able
Now let us suppose we want to compute the value x* such that the nonlinear objective
function F(x*) is a minimum. The basic notion of Newton's method for root finding is
to approximate the nonlinear constraint function c(x) by a simpler model (i.e., linear)
and then compute the root for the linear model. If we are to extend this philosophy to
optimization, we must construct an approximate model of the objective function. Just
as in the development of (1.5), let us approximate F(x) by the first three terms in a
Taylor series expansion about the current point x:

Notice that we cannot use a linear model for the objective because a linear function does
not have a finite minimum point. In contrast, a quadratic approximation to F(x) is the
6 CHAPTER 1. INTRODUCTION TO NONLINEAR PROGRAMMING

Figure 1.3: Minimization in one variable.

simplest approximation that does have a minimum. Now for a; to be a minimum of the
quadratic (1.18), we must have

Solving for the new point yields

The derivation has been motivated by minimizing F(x). Is this equivalent to solving
the slope condition F'(x) = 0? It would appear that the iterative optimization sequence
defined by (1.20) is the same as the iterative root-finding sequence denned by (1.6),
provided we replace c(x) by F'(x). Clearly, a quadratic model for the objective function
(1.18) produces a linear model for the slope F'(x). However, the condition F'(x) = 0
only defines a stationary point, which can be either a minimum, a maximum, or a point of
inflection. Apparently what is missing is information about the curvature of the function,
which would determine whether it is concave up. concave down, or neither.
Figure 1.3 illustrates a typical situation. In the illustration, there are two points
with zero slopes; however, there is only one minimum point. The minimum point is
distinguished from the maximum by the algebraic sign of the second derivative F"(x).
Formallv, we have
1.5. NEWTON'S METHOD IN SEVERAL VARIABLES 7

Necessary Conditions:

Sufficient Conditions:

Note that the sufficient conditions require that F"(x*} > 0, defining a strong local
minimizer in contrast to a weak local minimizer. which may have F"(x*) = 0. It is also
important to observe that these conditions define a local rather than a global minimizer.

1.5 Newton's Method in Several Variables

The preceding sections have addressed problems involving a single variable. In this
section, let us consider generalizing the discussion to functions of many variables. In
particular, let us consider how to find the n-vector XT — (xi xn) such that

For the present, let us assume that the number of constraints and variables is the same,
i.e., m = n. Just as in one variable, a linear approximation to the constraint functions
analogous to (1.5) is given by

where the Jacobian matrix G is denned by

By convention, the m rows of G correspond to constraints and the n columns to variables.

As in one variable, if we require that c(x) = 0 in (1.26). we can solve the linear system

for the search direction p, which leads to an iteration of the form

Thus, each Newton iteration requires a linear approximation to the nonlinear con-
straints c. followed by a step from x to the solution of the linearized constraints at x.
8 CHAPTER 1. INTRODUCTION TO NONLINEAR PROGRAMMING

Figure 1.4 illustrates a typical situation when n = m = 2. It is important to remark that

the multidimensional version of Newton's method shares all of the properties of its one-
dimensional counterpart. Specifically, the method is quadratically convergent provided
it is within a region of convergence, and it may diverge unless appropriate globalization
strategies are employed. Furthermore, in order to solve (1.28) it is necessary that the Ja-
cobian G be nonsingular, which is analogous to requiring that c'(x) ^ 0 in the univariate
case. And, finally, it is necessary to actually compute G, which can be costly.

Figure 1.4: Newton's method in two variables.

1.6 Unconstrained Optimization

Let us now consider the multidimensional unconstrained minimization problem. Sup-
pose we want to find the n-vector XT = ( x i , . . . , #„) such that the function F(x) =
F(XI, ..., z n ) is a minimum. Just as in the univariate case (1.18), let us approximate
F(x) by the first three terms in a Taylor series expansion about the point x:

The Taylor series expansion involves the n-dimensional gradient vector

1.6. UNCONSTRAINED OPTIMIZATION 9

and the symmetric n x n Hessian matrix

It is common to define the search direction p = x — x and then rewrite (1.30) as

The scalar term g T p is referred to as the directional derivative along p and the scalar
term pTHp is called the curvature or second directional derivative in the direction p.
It is instructive to examine the behavior of the series (1.33). First, let us suppose
that the expansion is about the minimum point x*. Now if x* is a local minimum, then
the objective function must be larger at all neighboring points, that is, F(x) > F(x*).
In order for this to be true, the slope in all directions must be zero, that is. (g*) p = 0.
which implies we must have

This is just the multidimensional analogue of the condition (1.21). Furthermore, if the
function curves up in all directions, the point x* is called a strong local minimum arid
the third term in the expansion (1.33) must be positive:

A matrix1 that satisfies this condition is said to be positive definite. If there are some
directions with zero curvature, i.e., p T H*p > 0, then H* is said to be positive semidef-
inite. If there are directions with both positive and negative curvature, the matrix is
called indefinite. In summary.
Necessary Conditions:

Sufficient Conditions:

The preceding discussion was motivated by an examination of the Taylor series about
the minimum point x*. Let us now consider the same quadratic model about an arbitrary
1
H* = H(x*) (not the conjugate transpose, as in some texts).
10 CHAPTER 1. INTRODUCTION TO NONLINEAR PROGRAMMING

point x. Then it makes sense to choose a new point x such that the gradient at x is zero.
The resulting linear approximation to the gradient is iust

which can be solved to yield the Newton search direction

Just as before, the Newton iteration is denned by (1.29). Since this iteration is based
on finding a zero of the gradient vector, there is no guarantee that the step will move
toward a local minimum rather than a stationary point or maximum. To preclude this,
we must insist that the step be downhill, which requires satisfying the so-called descent
condition

It is interesting to note that, if we use the Newton direction (1.41), the descent condition
becomes

which can only be true if the Hessian is positive definite, i.e., (1.35) holds.

1.7 Recursive Updates

Regardless of whether Newton's method is used for solving nonlinear equations, as in
Section 1.5, or for optimization, as described in Section 1.6, it is necessary to compute
derivative information. In particular, one must compute either the Jacobian matrix (1.27)
or the Hessian matrix (1.32). For many applications, this can be a costly computational
burden. Quasi-Newton methods attempt to construct this information recursively. A
brief overview of the most important recursive updates is included, although a more
complete discussion can be found in [46], [63], and [53].
The basic idea of a recursive update is to construct a new estimate of the Jacobian
or Hessian using information from previous iterates. Most well-known recursive updates
are of the form

where the new estimate B is computed from the old estimate B. Typically, this calcu-
lation involves a low-rank modification 7£(Ac, Ax) that can be computed from the pre-
vious step:

The usual way to construct the update is to insist that the secant condition
1.7. RECURSIVE UPDATES 11

hold and then construct an approximation B that is "close" to the previous estimate B.
In Section 1.3, the simplest form of this condition (1.15) led to the secant method. In
fact, the generalization of this formula, proposed in 1965 by C. G. Broyden [32], is

which is referred to as the secant or Broyden update. The recursive formula constructs
a rank-one modification that satisfies the secant, condition and minimizes the Frobenius
norm between the estimates.
When a quasi-Newton method is used to approximate the Hessian matrix, as re-
quired for minimization, one cannot simply replace Ac with Ag in the secant update.
In particular, the matrix B constructed using (1.48) is not symmetric. However, there
is a rank-one update that does maintain symmetry, known as the symmetric rank-one
(SRI) update:

where Ag = gfc — g f r ~ ] . While the SRI update does preserve symmetry, it does not
necessarily maintain a, positive definite approximation. In contrast, the update

is a rank-two positive definite secant update provided (Ax) Ag > 0 is enforced at

each iteration. This update was discovered independently by Broyden [33], Fletcher [52],
Goldfarb [65], and Shanno [98] in 1970 and is known as the BFGS update.
The effective computational implementation of a quasi-Newton update introduces a
number of additional considerations. When solving nonlinear equations, the search di-
rection from (1.28) is p = —G - 1 c. and for optimization problems the search direction
given by (1.41) is p = —H^ 1 g. Since the search direction calculation involves the ma-
trix inverse (either G"1 or HP 1 ), one apparent simplification is to apply the recursive
update directly to the inverse. In this case, the search direction can be computed sim-
ply by computing the matrix-vector product. This approach was proposed by Broyden
for nonlinear equations, but has been considerably less successful in practice than the
update given by (1.48), and is known as "Broyden's bad update." For unconstrained
minimization, let us make the substitutions Ax —> Ag, Ag —> Ax, and B —» B~ : in
(1.50). By computing the inverse of the resulting expression, one obtains

where

This so-called inverse positive definite secant update is referred to as the DFP update
for its discoverers Da.vidon [43] and Fletcher and Powell [55].
12 CHAPTER 1. INTRODUCTION TO NONLINEAR PROGRAMMING

Even though many recursive updates can be applied directly to the inverse matri-
ces, most practical implementations do not use this approach. When the matrices G
and/or H are singular, the inverse matrices do not exist. Consequently, it is usually
preferable to work directly with G and H. There are at least three issues that must be
addressed by an effective implementation, namely efficiency, numerical conditioning, and
storage. The solution of a dense linear system, such as (1.28) or (1.40), requires O(n3)
operations, compared with the O(n2) operations needed to compute the matrix-vector
product (1.41). However, this penalty can be avoided by implementing the update in
"factored form." For example, the (positive definite) Hessian matrix can be written in
terms of its Cholesky factorization as H — RR . Since the recursive update formulas
represent low-rank modifications to H, it is possible to derive low-rank modifications to
the factors R. By updating the matrices in factored form, the cost of computing the
search direction can be reduced to O(n2) operations, just as when the inverse is recurred
directly. Furthermore, when the matrix factorizations are available, it is also possible to
deal with rank deficiencies in a more direct fashion. Finally, when storage is an issue,
the matrix at iteration k can be represented as the sum of L quasi-Newton updates to
the original estimate B0 in the form

where the vectors U; and v» denote information from iteration i. If the initial estimate
BQ requires relatively little storage (e.g., is diagonal), then all operations involving the
matrix Bj. at iteration k can be performed without explicitly forming the (dense) matrix
Bfe. This technique, called a limited memory update, only requires storing the vectors u
and v over the previous L iterations.
We have motivated the use of a recursive update as a way to construct Jacobian
and/or Hessian information. However, we have not discussed how fast an iterative se-
quence will converge when the recursive update is used instead of the exact information.
All of the methods that use a recursive update exhibit superlinear convergence pro-
vided the matrices are nonsingular. In general, superlinear convergence is not as fast as
quadratic convergence. One way to measure the rate of convergence is to compare the
behavior of a Newton method and a quasi-Newton method on a quadratic function of n
variables. Newton's method will terminate in one step, assuming finite-precision arith-
metic errors are negligible. In contrast, a quasi-Newton method will terminate in at most
n steps, provided the steplength a is chosen at each iteration to minimize the value of
the objective function at the new point F(x) = F(x + ap). The process of adjusting a
is called a line search and will be discussed in Section 1.11.

1.8 Equality-Const rained Optimization

The preceding sections describe how Newton's method can be applied either to optimize
an objective function F(x) or to satisfy a set of constraints c(x) = 0. Suppose now that
we want to do both, that is, choose the variables x to minimize

subject to the m < n constraints

1.8. EQUALITY-CONSTRAINED OPTIMIZATION 13

The classical approach is to define the Lagranyian

where A is an m-vector of Lagrange multipliers.

In a manner analogous to the unconstrained case, optimality requires that derivatives
with respect to both x and A be zero. More precisely, necessary conditions for the point
(x*,A*) to be an optimum are

The gradient of L with respect to x is

and the gradient of L with respect to A is

Just as in the unconstrained case, these conditions do not distinguish between a point
that is a minimum, a maximum, or simply a stationary point. As before, we require
conditions on the curvature of the objective. Let us define the Hessian of the Lagmngian
to be

Then a sufficient condition is that

for any vector v in the constraint tangent space. If one compares (1.35) with (1.61),
an important difference emerges. For the unconstrained case, we require that the curva-
ture be positive in all directions p. However, (1.61) applies only to directions v in the
constraint tangent space.
Example 1.2. To fully appreciate the meaning of these conditions, consider the
simple example problem with two variables and one constraint illustrated in Figure 1.5.
Let us minimize

subject to the constraint,

The solution is at x* = (1,1). Now for this example, the Jacobian is just G =
Vc;T — (1,1), which is a vector orthogonal to the constraint. Consequently, if we choose
v = (—a, a) for some constant a ^ 0. the vector v is tangent to the constraint, which
14 CHAPTER 1. INTRODUCTION TO NONLINEAR PROGRAMMING

Figure 1.5: Equality-constrained example.

can be readily verified since Gv = (1)(—a) + (l)(a) = 0. At x*, the gradient is a linear
combination of the constraint gradients, i.e., (1.58) becomes

Furthermore, from (1.61), the curvature

is clearly positive.
Notice that, at the optimal solution, the gradient vector is orthogonal to the con-
straint surface, or equivalently that the projection of the gradient vector onto the con-
straint surface is zero. A second point is also illustrated in Figure 1.5, demonstrating
that the projection of the gradient is nonzero at the suboptimal point. Apparently then,
there is a matrix Z that projects the gradient onto the constraint surface. This implies
that an equivalent form of the necessary condition (1.56) is to require that the projected
gradient be zero, i.e.,

In a similar fashion, one can define an equivalent form of the sufficient condition (1.61)
in terms of the projected Hessian matrix

In other words, we require the projected Hessian matrix (1.63) to be positive definite.
Notice that the projection matrix Z has n rows and nj = n — m columns. We refer to
Other documents randomly have
different content
School of 42 St. Roch—a small figure
Perugino £6 10s.
Paduanino 43 Head of a duchess of Medicis—a
fragment £1 15s.
School of 44 Saints invoking Christ—two
Giotto illuminated miniatures £4 10s.
School of Titian 45 The Virgin and Child, on a
grassy bank, gathering flowers £3 15s.
School of 46 The Pentecost. A beautiful
Giotto miniature, on vellum, with rich
border £8 15s.
School of 47 The Crucifixion, with Saints. A
Giotto beautiful miniature, on vellum,
with rich arabesque border £4 14s. 6d.
School of Titian 48 Portrait of a Lady of the Medicis
Family: Pellegrina, daughter of
Bianca Capello, in a richly
ornamented dress £4 14s. 6d.
A. del Sarto 49 The Resurrection. An interesting
small work in the Master's first
manner. From the de Angelis
Gallery, at Sienna £4 14s.
C. Maratti 50 The Holy Family £3 3s.
Bronzino 51 The Virgin and Child, with St.
Joseph and St. John. A very
grand and beautiful design—
circular, on panel [£21 0s. 0d.]
G. da Carpi 52 The Virgin, in a crimson and
blue dress, seated, with the
Infant in her lap, before a
sculptured portico; a green
drapery suspended above—
circle on panel. The Orsini Arms
on the frame £141 15s.
Luigi Agresti 53 The Last Supper. Very richly
coloured, with the engraving £6
School of 54 The Virgin and Child, enthroned,
Brescia with saints and angels in
adoration £4
F. Vanni 55 The repose of the Holy Family—
small £2
Schedone 56 The Virgin and Child £1 13s.
School of 57 A female saint, holding a salver
Parma of fruit—very elegant £4 4s.
School of 58 The Holy Family, with St. Francis
Ferrara and St. Jerome £9 9s.
Guardi 59 A view on a canal, at Venice,
with figures £11 0s. 6d.
Guardi 60 A view on the grand canal—the
companion £11 0s. 6d.
Scorza, of 61 A pastoral landscape
Genoa £1 6s.
S. Rosa 62 A romantic bay scene, with
figures—evening £4 10s.
Serani 63 St. Cecilia, playing on the viol
da gamba £5 7s. 6d.
Testaferrata 64 A Roman piper £1 1s.
Antonilez di 65 St. Raymond of Penaforte. From
Serabia the Standish Gallery £3 13s. 6d.
Montelinez di 66 St. Anthony, seated reading,
Serabia near a chapel, with
mountainous background. From
the Standish Gallery £2 2s.
Zurbaran 67 The Madonna of Mercy—four
figures kneeling round her. From
the Standish Gallery [£2 15s.]
Murillo 68 The vision of St. Augustine of £199 10s.
Canterbury: the saint is washing
 the feet of the Saviour, who
appears in the likeness of a
pilgrim; from his mouth proceed
the words "Magne Pater
Augustine tibi commendo
Ecclesiam meam." This fine
gallery picture was purchased
from Don Julian Williams, by Mr.
Standish, for £600, at Seville, in
1825; it was originally painted
for the nuns of San Leandro
Order of St. Austen, and sold by
them during the troubles caused
by the army of Soult, in 1810,
to Dr. Manuel Real, from whom
it passed to Don J. Williams.
The picture is mentioned in the
work of Herrera and d'Aviles
Guia de Seville, 1832
Velazquez 69 Portrait of a Cardinal, seated,
holding a book, the chair
surmounted by shields of arms.
Full of dignified character. From
Cardinal Fesch's gallery £4 4s.
School of 70 La Madonna Addolorata, in a
Cologne crimson dress, a light-coloured
robe. A very dignified figure. A
fragment [£1 10s.]
Wilhelm, of 71 The Marriage of St. Catherine
Cologne with St. Agnes. They are in the
foreground of a landscape, with
buildings in the distance. From
M. Wyer, of Cologne £14 3s. 6d.
Van der Maire 72 St. Catherine, presenting a £5 15s. 6d.
devotee. An interesting
 fragment
Van Eyck 73 A fine dyptic, with the
Annunciation: the Virgin
kneeling, the Angel in a rich
dress, holding a sceptre:—the
portrait of the donor outside.
From the Collection of M. Wyer,
of Cologne £39 18s.
Henri Blaes La 74 A tryptic, with the Virgin and
Civetta Child in the centre, seated, in a
landscape; St. Christopher and
St. Anthony on the wings; an
owl, the emblem of the Master.
From the Collection of M. Wyer,
of Cologne £17 6s. 6d.
Dionysius 75 The Crucifixion: The Virgin and
Calcar St. John weeping, with
landscape background £8 8s.
School of 76 St. Natalitia, seated, holding a
Hemmelinck book, on which is a hand, cut
off; with architectural
background. From the same
collection £12 1s. 6d.
Matth. 77 Portrait of Philip le Bel, in a
Guinendenwald crimson dress and black hat,
wearing the collar of the Golden
Fleece £6.
Van der Goes 78 The Virgin and Child,
enthroned; a damask drapery
behind; landscape background
seen on each side £22 1s.
Lucas van 79 A very small female head—a
Leyden fragment 18s.
Martin Schoen 80 A tryptic: the Crucifixion, with £12 1s. 6d.
the figures carved in wood, and
 painted background in the
centre; the wings painted with
the six stations; carved canopy
work over the centre; the
descent from the Cross painted
on the outside
Sustermans 81 Portrait of Galileo £4 10s.
Sustermans 82 Portrait of a Florentine lady £2 4s.
Van Dyck 83 The Adoration of the Magi—a
sketch in grisaille £2 8s.
Van Dyck 84 Portrait of the Earl of Strafford,
in a black dress. Purchased from
the Earl of Mar's collection, in
1805 £5 5s.
Teniers 85 A landscape, with peasants and
poultry near a cottage—upright
—on copper £3 7s. 6d.
Camphuyzen 86 A farm, with cattle, and a man
milking a cow near a well. Very
richly coloured £6 10s.
Jan Steen 87 Portrait of a Burgomaster £5 15s.
Poelemberg 88 The Riposo of the Holy Family,
under a ruined building £2 10s.
Wouvermans 89 Travellers, reposing under a
sunny bank, near a pool of
water £33 12s.
Van Falens 90 Camp suttlers, with horsemen
and numerous figures. From the
Collection of Sir James Stuart £5 10s.
Swaneveldt 91 A study of ruins—on paper £1 2s.
Watteau 92 A fête champêtre £1 10s.
Rigaud 93 Portrait of a French lady,
holding a row of pearls £3 10s.
Venetian 94 Portrait of the admirable
Crichton, in black dress, seated
holding a sword and a book;
with long inscription. Dated
1581, with the engraving £13 2s. 6d.
Roman School 95 Portrait of the Cardinal of York,
in his robes. Purchased at his
villa, at Frascati £1 10s.
Sir P. Lely 96 Portrait of the Countess of
Southesk, (la belle Hamilton) in
a white satin dress, seated,
holding a viol da gamba, in a
landscape, from the collection
of C.R. Sharpe, Esq. [£7 17s. 6d.]
Sir Joshua 97 A very small head of a lady
Reynolds 15s.
Anthony 98 Henselope Burn 14s.
Andrew Wilson 99 The Cascatelle, at Tivoli, with
shepherds and goats in the
foreground, admirably painted £33 12s.
J.M.W. Turner, 100 A farm in the Highlands
R.A. £2 8s.
Rev. J. 101 The Trosacks. A beautiful
Thompson finished study, given by the
artist to Mr. Dennistoun in 1829 £3
J.M.W. Turner, 102 Fishing boats caught in a squall
R.A. £8 15s.
Millais, A.R.A. 103 A cottage barn, in Essex: a
sketch of figures on the back £4 10s.

WORKS OF MEDIÆVAL ART, AND CURIOSITIES

104 Eleven silver touch-pieces, for the King's Evil, of
the Stuarts; and three bronze Papal coins £5
105 A pair of red silk stockings, worked with gold.
Belonged to the last Duke of Urbino £6
106 A curious ivory die, representing a man seated;
and four silver dice, in the form of men and
women seated £1
107 A pair of brass church candlesticks £1 15s.
108 The Virgin and Child—a relief, in bronze £1 15s.
109 The Flagellation—a relief, in bronze £1 15s.
110 A miniature portrait of Queen Mary, mounted in
silver, with slab of agate on the back £4 6s.
111 Raffaele School—Lo Spasimo di Sicilia—a
drawing, in Indian ink and pen 6s.
112 A chalice, of silver, and copper gilt, with three
busts of Niello work on the base £3 15s.
113 A female saint, in embroidery 5s.
114 Head of St. Peter, in tapestry. From the Cardinal
of York's Villa £3 6s.
115 The Crucifixion, worked in ancient lace for an
altar cloth £2 2s.
116 A very rare caterpillar's web, of unusual size [£2]
117 St. Mary, of Egypt, of pietra-dura, on lapis-lazuli
ground £12 10s.
118 A Majolica plate, with St. Jerome, in a landscape:
signed by Maestro Giorgio, 1521—imperfect £3 15s.
119 A fragment of a Majolica plate, with Mercury,
with the initials of Maestro Giorgio, 1534 15s.
120 The agony in Gethsemane—a Limosine enamel £1
121 St. Dietburgha—painted on a caterpillar's web £1 6s.
122 A half dyptic, with two saints in relief in ivory,
and Byzantine inscription £1 11s.
123 A crucifix, elaborately carved in boxwood,
containing a rosary of silver thread £1
124 A large bronze Papal seal, with the Holy Family,
and 8 smaller bronze seals—one of them, Johann
Russell £2 5s.
125 Venus on a dolphin—a Venetian bronze, on
oriental alabaster plinth £3 10s.
126 A bronze inkstand, supported on eagles, and
surmounted by a figure £3 10s.
127 The Entombment—a relief, in bronze £10 10s.
128 A small ivory dyptic, with the Crucifixion, and the
Virgin and Child, with two saints, in high relief, on
gothic arches £10 10s.
129 A very interesting Pax, of Niello work, with Christ
bearing his Cross, and appearing to Mary,
inscribed above "Jacobus Suannes Cole"; the
dead Christ, and emblems of the Crucifixion, in
the lunette above £10 10s.
130 Another Pax, of niello, with the dead Christ and
angels, inscribed beneath, "Pax tibi Pilastus," and
frieze of arabesque; the Creation above—
mounted in ivory £9
131 A curious bone tryptic, with the Crucifixion,
attended by saints; St. Peter and St. Paul on the
wings £8 18s. 6d.
132 A very interesting early Pax, of Niello, with the
Virgin and Child enthroned, the latter holding a
rosary; two saints kneeling on each side; a die on
the ground in the centre [£2 15s.]

The total amount realised at the sale was £1398 15s. 6d.
 AUTHOR’S PREFACE
(1851)

D
URING nearly one hundred and ninety years, five Dukes of
Urbino well and ably discharged the duties of their station,
comparatively exempt from the personal immoralities of their
age. The rugged frontier of their highland fief had, in that time,
been extended far into the fertile March of Ancona, until it embraced
a compact and influential state. Saving their subjects, by a gentle
and judicious sway, from the wild ferments that distracted
democratic communities, and from the yet more dire revolutions
which from time to time convulsed adjoining principalities, they so
cultivated the arts of war, and so encouraged the pursuits of peace,
that their mountain-land gained a European reputation as the best
nursery of arms, their capital as the favoured asylum of letters. That
glory has now become faint; for the writers by whom it has been
chiefly transmitted belong not to the existing generation, and
command few sympathies in our times. But the echoes of its fame
still linger around the mist-clad peaks of Umbria, and in the
dilapidated palace-halls of the olden race. To gather its evanescent
substance in a form not uninteresting to English readers, is the
object of the present attempt. Should it be so far successful as to
attract some of his countrymen to the history, literature, and arts of
Italy, they will not, perhaps, be ungrateful to the humble pioneer
who has indicated a path to literary treasures hitherto inadequately
known to them. For such an undertaking he possesses no
qualification, beyond a sincere interest in the past ages of that sunny
land, and a warm admiration for her arts during their epoch of
brilliancy. But a residence there of six years has afforded him
considerable opportunities of collecting materials for this work, which
he has been anxious not to neglect.
A great portion of the duchy of Urbino, including its principal towns,
has been thrice visited, and nearly every accessible library of Central
Italy has been examined for unedited matter. To these researches,
time and labour have been freely given; and in the few instances
when his attempts were foiled by jealousy or accident, the author
has generally had the satisfaction of believing that success would
have been comparatively unproductive. To this, two exceptions
should be mentioned. He was prevented by illness from recently
visiting the libraries or archives at Venice; and the Barberini Library
at Rome has been entirely closed for some years, in consequence of
a disgraceful pillage of its treasures. Should the latter be again made
accessible, the MSS. amassed by the Pontiff under whom Urbino
devolved to the Church, and by his nephews, its two first Legates,
can hardly fail to throw much light upon the duchy. The invaluable
treasures of the Vatican archives have been to him, as to others, a
sealed book; but the Urbino MSS. in the Vatican Library, those of the
Oliveriana at Pesaro, and of the Magliabechiana at Florence, have
afforded copious sources of original information, and have supplied
means for rectifying omissions and errors of previous writers. Some
of these materials had been freely drawn upon by Muzio, Leoni, and
Baldi, biographers of the early dukes of Urbino, who have not,
however, by any means exhausted the soil; the amount that
remained for after inquirers may be estimated from the single
instance of Sanzi's almost unnoticed rhyming Chronicle of Duke
Federigo, in about 26,000 lines.
The reigns of Dukes Federigo, Guidobaldo I., and Francesco Maria I.,
from 1443 to 1538, formed the brightest era of Urbino, and included
the most stirring period of Italian history, the golden age of Italian
art; but our regnal series would be incomplete without Dukes
Guidobaldo II. and Francesco Maria II., who prolonged the
independence of the duchy until 1631, when it lapsed to the Holy
See. Its history thus naturally divides itself into five books,
representing as many reigns; yet, as these sovereigns were of two
different dynasties, it will be convenient to consider separately the
origin of each, and the influence which they respectively exercised
on literature and the fine arts, thus giving matter for four additional
books. In Book First of these we shall briefly sketch the early
condition of the duchy, with the establishment of the family of
Montefeltro as Counts, and eventually as Dukes, of Urbino; but,
regarding Duke Federigo as the earliest of them worthy of detailed
illustration, we shall, in Book Second, with his succession, enter
upon the immediate scope of our work.

Among many interesting publications upon Italy which have recently

issued from the English press, is that of Signor Mariotti.[3] With a
command of our language rarely attained by foreigners, he has
clothed a vast mass of information in an exuberant style, savouring
of the sweet South. As an episode to his sketch of Tasso, he
dedicates to the two dynasties who ruled in Urbino a single page, in
which there occur seven misstatements. John or Giovanni della
Rovere was never sovereign of Camerino; his cousin, Girolamo
Riario, held no ecclesiastical dignity; the "unrivalled splendour" of
the Montefeltrian reign at Urbino did not extend over even one
century; the wife of Giovanni della Rovere was neither daughter nor
heiress of Guidobaldo I. of Urbino, nor had she any "just claim to his
throne"; Duke Francesco Maria did not remove either his library or
treasures of art to Mantua. These slips, by a writer generally
painstaking and correct, surely indicate some deficiency in the
accessible sources of information regarding a principality which has
for centuries been proverbial, in the words of Tasso, as "the stay and
refuge of gifted men."
The truth is, that although the Dukes of Urbino figure everywhere as
friends of learning and patrons of art, no work has yet appeared
establishing their especial claim to such distinction, in a land where
courts abounded and dilettanteship was a fashion. That of Riposati
has indeed given us the series of these sovereigns, but his
biographical sketches are meagre, and chiefly illustrative of their
coinage. The lives of Dukes Federigo and Francesco Maria I., by
Muzio and Leoni, are excessively rare; Baldi's crude biographies are
either recently and obscurely published, or remain in manuscript.
Out of Italy these authors are scarcely known. This paucity of
illustration is not, however, the only cause why these princes have
continued in unmerited obscurity. Whilst endeavouring to guard
himself against undue hero-worship, and to subject the policy and
character of those sovereigns to the tests within his reach, the
author has been obliged in some instances to assume the functions
of an advocate, and to defend them from charges unjustly or
inadvisedly brought. This will be especially found in the life of Duke
Francesco Maria I., who, as the victim of Leo X., and the opponent of
Florence, has met with scanty justice from the three standard
historians of that age in Italy, France, and England. The patriotism of
Guicciardini, as a Florentine, was inherently provincial; as a partisan
of the Medici, he had no sympathies with a prince whom they hated
with the loathing of ingratitude; as an annalist he never forgot the
day when he had cowered before the lofty spirit at the council-
board. All that he has written of Francesco Maria is therefore tinged
with gall, and his authority has been too implicitly followed by
Sismondi, who, uniformly biassed against princes by his democratic
prejudices, and seeing in Guicciardini an eminent denizen of a
nominal republic, and in the Duke a petty autocrat, decided their
respective merits accordingly. Again, Roscoe could save the
consistency and justice of Leo only by misrepresenting the character
of his early friend and eventual victim, and has not shrunk from the
sacrifice. It has thus happened that, whilst ordinary readers have
scanty access to details regarding Urbino and its dynasties, these
names have been unduly excluded from many a page in Italian
annals which they were well qualified to adorn.[4]
To separate from the tangled web of Italian story threads of local
and individual interest would be fatal to unity of texture and subject.
It will, therefore, be necessary to treat Urbino and its Dukes as
integral portions of the Ausonian community, and, while
distinguishing every characteristic detail, to view them as subsidiary
to the general current of events. But, since this course offers at
every moment temptations to launch our tiny bark on a stream
perilous to its pilot, prudence will keep us mostly among those
eddies which, unheeded by more skilful mariners, may afford leisure
for minute observation. If it be thought that the martial renown of
Federigo and Francesco Maria I. merited more ample accounts of
their campaigns, we may plead that arms are but a portion of our
object. To mankind battle-fields are instructive chiefly from their
results; while foreign and domestic policy, the progress of civilisation
and manners, of letters and art, are in every respect themes of
profitable inquiry.
In a work undertaken with the hope of attracting general readers to
the history and arts of Italy, controversial disquisitions would be
misplaced. The student may detect occasional attempts to reconcile
contradictory narratives and jarring conclusions; but religious
discussion is excluded from these pages. The author is a Protestant
by birth and by conviction, but it has been his endeavour to judge
with candour, and speak with respect, of a Church which is the
"parent of our religion," and which, during a great portion of his
narrative, was catholic in the strict sense of that often misapplied
term. He has mentioned without flattery, extenuation, or malice,
such private virtues and vices of the various pontiffs as fell within
the scope of his inquiry, leaving it to others to fix the delicate line
which is supposed to divide personal errors from papal infallibility.

A considerable portion of these volumes was written in Italy, before

the close of Pope Gregory's reign, and under impressions formed
upon the existing state of the country. It has been their author's
good fortune to know much of that attractive land during the last
twenty years of the long peace, and to admire her substantial
prosperity and steady progress. Between 1825 and 1846 he has
seen in her cities new streets and squares rising, thoroughfares
opened, gas-lights generally introduced, ruinous houses substantially
rebuilt, crumbling churches and palaces renovated, shops enlarged
and beautified, cafés, hotels, and baths multiplied and decorated,
public drives and gardens created, equipages rivalling those of
northern capitals, museums formed, galleries enriched, the dress
and comforts of the population greatly improved, the street
nuisances of Rome removed, the lazzaroni of Naples clothed.
In the rural districts he has observed cultivation spreading, waste
lands reclaimed, irrigation and drainage carried on, the great
highways rendered excellent, whole provinces opened up by new
roads, railways rapidly extending, rivers and torrents bridged,
palatial villas springing up round the towns and watering-places,
banditti suppressed, the peasantry ameliorated in aspect. He has
learnt, from crowded ports and spreading factories, that capital was
increasing and industry being developed.
He has also noticed that, without organic changes, the political
condition of the people was being modified; that Tuscany enjoyed
the mildest of paternal governments; that in Lombardy, Piedmont,
and Naples, many repressive statutes were in abeyance; that in
Turin and Florence restraints upon the press were tacitly being
relaxed; that scientific congresses were generally permitted, and
political economy freely discussed; whilst, in regard to Rome, he
ascertained the practical truth of a popular sarcasm, that prohibitory
laws were usually binding but for three days.
While conscious of all this progress, the author felt that much
remained to be done. He knew that the advance of the country was
only comparative, and rendered more apparent by her long previous
stagnation. He daily had before him solecisms in policy, errors of
administration, official indolence or corruption; above all, ample
proofs that priests were no longer adapted for ministers of state. He
believed that intellect was needlessly or unwisely shackled, and that,
to ardent or speculative minds, the full blaze of knowledge might be
less deceptive than a compulsory twilight.
But, on the other hand, he was deeply convinced that, in material
welfare, the Italian people were already far above the average; that
any sudden change was more likely to endanger than to augment it;
that, to a nation so listless yet so impressionable, so credulous but
so suspicious, self-government was a questionable boon; at all
events, that the mass of its present generation was infinitely too
ignorant and unpractised, possibly too conceited and self-seeking, to
comprehend the theory of a constitution, or to perform the duties it
would necessarily impose. He knew further, that those who vaguely
longed for change were usually blind to the benefits which their
country already enjoyed, and had no definite or plausible plan for
the removal of its grievances without perilling its advantages. He felt
satisfied that, should an occasion ever present itself for testing their
Utopian theories, native leaders, united in aims and worthy of their
reliance, would be wanting. The movement party in Italy then
scarcely numbered a man who had a considerable property to stake,
a social position to lend him influence, or tried business habits to
gain the confidence of his fellow-citizens. Those who stood prepared
to pilot the vessel through revolutionary storms were, for the most
part, persons whose detected intrigues, or rash outbreaks, had
already driven them, with little credit, into exile, where, cut off from
intercourse with home, and associating chiefly with kindred spirits
expelled from other lands, they forgot much which it was important
to keep in view, and learned little of that candour and moderation
which are the true leaven of politics. Neglecting there those practical
reforms of which Italy stood really in need, they devoted themselves
to one idea. They set up the phantom of political unity as a new
faith; they decreed that its worship should be the condition of their
country's resurrection, and that all who demurred to it should be
hunted down. Had they read Dante, or remembered what they
hourly had seen, heard, and said in their native land, they would
have known that their idol, like the image in Nebuchadnezzar's
dream,[5] was of incongruous and incompatible materials; that their
unitarian scheme was antipathic to every passion and prejudice of
those upon whom they would thrust it.
Under such impressions were written the very few allusions to the
actual state of Italy which this work contains. The aspirations of her
regenerators after nationality and constitutional freedom have since
been fostered by her spiritual ruler, and prematurely fired by an
explosion of French democracy. Subsequent events, under altered
circumstances, may accordingly seem to have invalidated opinions
therein expressed; but the end is not yet. The present continues
overshadowed by gloom, and the torch of hope glimmers but dimly
in the distance. A sincere interest in the country and its people
dictates our prayer that the God of nations may grant an issue
realising the fondest anticipations of genuine patriotism, and
eventually crown these struggles with results compensating their
recent evils.
Yet when we recollect the condition of Italy as we left her shores
four short years ago,—when we contrast the calm then around her
institutions, the stillness of her every-day life, the careless ease of
her nobles, the physical enjoyment of her middle classes, the simple
well-being of the peasantry under their own vines and fig-trees,—we
must sigh to see so much positive happiness perilled for contingent
ameliorations which, if ever attained, may, like most political
experiments, fail to realise the promised benefits.

"Let him who sees mad war like deluge sweep

Surrounding regions, learn his peace to prize;
Let the poor bark with sides unripped, which tries
In vain by helm and sail its course to keep,
Make for the port. He lives perchance to weep,
Who quits the genial air and smiling skies
For depths unknown. O blind desire unwise
Of mortals, spurning thus on earth to creep!
O when, in this his mouldering garment frail,
Did man, whose thread soon breaks and joins no more,
Clear his own path, or by his power prevail?"[6]
In a work of history, party politics ought to have no place; and when
the nations are moved there is little inducement to assume a
prophet's mantle. We, therefore, gladly leave a topic on which
perhaps too much has been said. Possibly some Italians, to whom
we have formerly represented that it were

"Better to bear the ills we know,

Than rush on others that we wot not of,"

may yet admit the truth of this suggestion. May they never
personally realise the adage, that those who originate revolutions
reap all their evils, without living to share their fruits!

A few words regarding the method adopted in these volumes. Of the

names most conspicuous in Italian literature and art, a considerable
proportion will there find a place; but readers who expect to see
their productions enumerated, and their merits submitted to
exhaustive criticism, will be disappointed. All that our limits permit,
after rapidly sketching the revival of knowledge and the progress of
that sacred painting which emanated from Umbria, is to mention
those who have contributed to shed lustre over the duchy of Urbino,
or who shared the patronage of its princes. The amount of notice
allotted to each is therefore proportioned rather to its local
importance than its absolute excellence; but, satisfied from
experience how seldom a wide-spread interest attaches to individual
details, our aim has ever been to generalise even those points
demanding a more specific notice in connection with our immediate
subject.
As the recurrence of foot-notes in a popular narrative unpleasantly
distracts the reader from its continuous course, these have been
avoided, unless when especially called for; and the necessity for
them in citing references has been in a great degree anticipated, by
prefixing a list of the leading authorities consulted, which it is hoped
will generally bear out views that have been honestly formed, after
examining what seemed the best sources of information. Extracts
have been introduced, where it appeared desirable to preserve the
style or words of an author; but they are in most cases rendered
(literally rather than with elegance) into English, except such
specimens of poetry as could not be fairly estimated from a
translation. Documents and episodical details, which would have
encumbered the text, are appended to the respective volumes.[7]
The majority of proper names being Italian, are written in that
language, excepting such as, like those of places, and titles of popes
and sovereigns, have long been familiar to English ears in a different
orthography. In such matters uniformity of practice is the main
object to be attended to, and having to choose between names as
they were actually used and their English synonyms, we have
preferred Giacomo Piccinino, Giulio Romano, and Lorenzo de' Medici,
to James the Little Fellow, Julius the Roman, and Lawrence of the
Medici.[8] There will often be mentioned districts and divisions of
Italy which are defined by no exact political or geographical limits; it
may therefore be well here to explain in what sense these somewhat
convertible terms are employed. Central Italy may be considered to
contain the papal territory and the three Tuscan duchies; Upper and
Lower Italy include all the Peninsula, respectively to the north and
the south of these states. Again, Lombardy is used as a generic term
for the whole basin of the Po, the Polesine being that portion of its
delta, north of the river, which belonged to the Dukes of Ferrara.
Romagna stretches from the Po to the Metauro, from the Apennines to
the Adriatic; La Marca, or the March of Ancona, continues the same
sea-board to the Tronto: these two districts were long the cradle of
Italian prowess, the allotment-land of petty princes; both were
partially comprehended within the more ancient landmarks of Umbria,
a mountain province lying east of the Tiber. The lower basin of this
classic stream contained Sabina on the east, and the Patrimony of St.
Peter on the west; the Comarca lying south of the Teverone stream,
and the whole wide plain around Rome being called the Campagna.
Tuscany, including the Sienese, ran northwards from the Patrimony,
beginning below Orbetello; and Naples is familiarly called by Italians
The Kingdom, having, until a recent date, been the only royal state in
their fatherland.
Our chronology also requires the use of certain conventional terms,
which ought to be defined. Assuming the close of the fifteenth
century as the zenith of Italy's glory in letters and arts, in politics
and arms, the only word specifically indicating that period is cinque-
cento; but seeing that its lustre was attained under military and civil
institutions, and was rendered permanent by studies and artistic
creations, derived from the middle ages and breathing their spirit,
the phrase mediæval is extended to include that period.

Few things are more baffling to students of history than the true
worth of money in different states and ages, and its relative value in
reference to our own standards. It is impossible to over-estimate the
convenience which tables, showing the fluctuations of currency and
prices among different nations, would afford; but the difficulties of
completing them may perhaps be insuperable. In order to supply
this desideratum, however imperfectly, a few observations are here
submitted.
In considering the value of money at different periods, a variety of
circumstances must be kept in view. There are, however, four
elements to be embraced by all calculations for such a purpose: (1)
the comparative weight of the coinage; (2) the respective amounts
of alloy introduced into the standard of precious metals; (3) the
effect produced on gold and silver value by the discovery of
America; (4) the fluctuations in prices of commodities. The last of
these elements includes and depends upon the others, so that a
tariff of prices at various times might be practically sufficient for the
object contemplated. The impediments, however, to obtaining such a
tariff are apparently insurmountable. Statistical facts, incidentally
mentioned by historians, or gleaned from original documents, must
be received with large allowance. Articles of costly luxury in one age
became abundant in another, and are at all times affected by local or
temporary causes. Quality was also variable; horses, oxen, sheep,
and poultry, reared or fed in rude times or uncultivated districts,
cannot fairly be compared with those perfected by care and
expense; the same may be said of wines, fruits, clothing; even land
is saleable according to its condition, fertility, or situation. The test
usually resorted to in such inquiries is corn; but weights and
measures, seldom uniform, are with difficulty ascertained at remote
periods, while exceptional prices are more frequently noted than
average ones, by observers prone to record striking events rather
than every-day facts. There are, however, some apparently admitted
data not altogether unavailable for our immediate purpose.
During the period embraced by our memoirs of Urbino, the standard
of value prevalent in most parts of Italy was the golden florin or
ducat. Of these probably equivalent terms, the former was generally
employed in Central Italy, the latter in Lombardy. According to
Villani, the florin of Florence, in 1340, weighed 72 grains of pure
gold, 24 carats fine. Sismondi, in referring to a period about a
century later, estimates its weight at ⅛ of an ounce, or 60 grains.
Orsini reckons it, in 1533, at 70 grains, 22 carats fine. On the whole,
it appears, from Cibrario and other authorities, that this coin, and its
successor the zecchino, have maintained an almost uniform weight
down to the present time. Assuming that gold in Italy had then the
same coinage-value as in England, it appears from calculations,
founded upon Fleetwood's data, that the florin was, at these various
periods, equivalent in contemporary English coin to 3s. 6d., 4s. 8d.,
and 5s. 10d. Again, the ducat of Venice is estimated by Daru at 4
franks in 1465, at 4⅓ in 1490, and by Sanuto at 4s. English in 1500.
Riposati, in a careful analysis of the coinage of Gubbio, proves that
the conventional Urbino florin of 1450 should have contained
63434/59 grains of silver, besides alloy, which would at that time have
yielded 3s. 9d. English, or at our present pure silver value (5s. 6d. to
the ounce) 7s. 3¼d.
It would follow, from these several opinions, that the florin or ducat
of Italy, in the fifteenth century, was equal to from 3s. 6d. to 4s. 6d.
in contemporary English circulation, which disposes of two elements
for our calculation. The remaining two must have been inadequately
kept in view by Cibrario, Ricotta, and Audin, who respectively value
the florin of 1400 as now worth 16⅔ francs, that of 1490 at 14
francs, and that of 1500 at 12 francs; while in the Library of
Entertaining Knowledge it is set down at 10s. English in 1480. But if
we assume the analogy of English prices as collected by Fleetwood,
the result will be very different. From these it appears that an
average cost of wheat and oats per quarter, in the fifteenth century,
was about 5s. 2d. and 2s. 6d., while the wages of labourers and
artisans were respectively 3½d. and 4½d. a day. Accordingly, if corn
be taken as the test, money was then ten times beyond its modern
value; while, if we include labour and luxuries, the actual
depreciation must appear much greater. We are greatly encouraged
to find such an inference not very different from that adopted by
three recent and important authorities. Prescott values the Spanish
ducat of 1490 at 39s. 4d., and Macaulay states that of Florence in
1340 at 40s. sterling, while Sismondi calculates it at about 48 francs.
On the whole, then, we venture to assume that the Italian ducat or
florin of the fifteenth century was nearly equal to the present
Spanish dollar, and that it would have purchased about twelve times
the amount of necessaries and luxuries which that coin now
represents in England—a discrepancy of course lessened in the next
and each succeeding age, especially as the precious metals
continued to flow in from the new hemisphere. This estimate is,
however, offered with great deference, and only as a general
approximation to the truth, by no means applicable to numerous
exceptional cases.[9]

In closing these preliminary observations, it is a pleasing duty to

acknowledge the facilities obligingly placed at the author's disposal
by kind friends in Italy and at home. The urbanity with which
Monsignore Laureani afforded every assistance compatible with the
stringent regulations of the Vatican Library, demands a tribute
tempered by regret that death should have prematurely removed
him from a trust which he usefully and gracefully discharged. To Don
Pietro Raffaele, of the Oliveriana Library at Pesaro, and to the Abbé
Francesco Raffaele Valenti, of the Albani Library at Urbino; to Signor
Luigi Bonfatti, of Gubbio; to the archivists of many towns, and to the
directors of not a few galleries in Italy, a large debt of gratitude has
been incurred. The intimate acquaintance with the treasures of
Italian art possessed by the Commendatore Kestner, minister from
the Court of Hanover at the Holy See, was, with his wonted kindness
and courtesy, freely rendered available. Mr. Rawdon Brown, whose
profound knowledge of Venetian history and antiquities will, it is
hoped, be ere long appreciated in England, as it already is in the
Lagoons, has communicated most important documents, which the
author was unable personally to inspect. Mr. F.C. Brooke, of Ufford
Place, Suffolk, has likewise supplied some valuable notices. The
embellishments of these volumes owe much to the friendly
assistance of Mr. Lewis Gruner, an artist whose generous character
and happy exemption from professional jealousies are not less
remarkable than the success of his burin and the excellence of his
taste. With a liberality unusual among English collectors, Dr.
Wellesley, Principal of New College Hall, Oxford, threw open his
stores of Italian historic art, and allowed the use of several rare
medallions. To these, and to many whose good wishes have cheered
him on, the author's thanks are thus heartily, though inadequately,
offered.
 CONTENTS
PAGE
Introduction vii
Memoir of the Author xiii
Catalogue of the Picture Sale of the Author xix
Author’s Preface xxix
List of Illustrations of Volume I. xlix
Chronological Table of Contents of Volume I. li

BOOK FIRST
OF URBINO AND ITS EARLY COUNTS

CHAPTER I
Topography of the Duchy of Urbino—Origin of the Italian
communities—Their civil institutions and military system—Their
principle of liberty—Political divisions of Romagna; opposed to
modern speculations regarding centralization 3
CHAPTER II
Origin of the Counts of Montefeltro, and of their sovereignty in
Urbino and the surrounding country—Their early genealogy—
Guido Count of Urbino—Antonio Count of Urbino 22
CHAPTER III
Guidantonio Count of Urbino—The Ubaldini—Oddantonio Count
of Urbino—Is made Duke—His dissolute habits and speedy
assassination 42
 BOOK SECOND
OF FEDERIGO DI MONTEFELTRO, COUNT AND
SECOND DUKE OF URBINO

CHAPTER IV
The birth of Count Federigo—Condition of Italy—His marriage
and early military service—The Malatesta his inveterate foes—
He takes S. Leo—Is invested with Mercatello 61
CHAPTER V
Count Federigo succeeds to Urbino and acquires Fossombrone
—His connection with the Sforza family, whereby he incurs
excommunication—His campaign in the Maremma—Loses his
eye in a tournament 85
CHAPTER VI
Count Federigo enters the Neapolitan service—His two
campaigns in Tuscany—Fall of Constantinople—Peace of Lodi—
Nicholas V.—The Count's fruitless attempt at reconciliation with
Sigismondo Pandolfo Malatesta, followed by new feuds with him
—Death of his Countess Gentile 102
CHAPTER VII
Count Federigo's domestic life—His second marriage—New war
for the Angevine succession to Naples—Battle of San Fabbiano
—Conclusion of the war—Humiliation of the Malatesta 120
CHAPTER VIII
Count Federigo's home administration and court—Description of
his palace and library at Urbino—His other palaces—The
resources of his state 147
CHAPTER IX
Count Federigo's varied engagements—Battle of La Molinella—
Death and character of his enemy Malatesta—Affairs of Rimini 177
CHAPTER X
Birth of Prince Guidobaldo—Count Federigo captures Volterra—
Is again widowed—Receives the Garter and the Ermine—Is
made Duke of Urbino—His patronage of learned men 207
CHAPTER XI
The Duke of Milan assassinated—Count Girolamo Riario—The
Pazzi conspiracy—Duke Federigo's campaigns in Tuscany—
Progress of the Turks 233
CHAPTER XII
The war of Ferrara, and the death of Duke Federigo—His
258
character and portraits

BOOK THIRD
OF GUIDOBALDO DI MONTEFELTRO, THIRD DUKE OF URBINO

CHAPTER XIII
The early promise of Duke Guidobaldo I.—Count Girolamo
Riario assassinated—The Duke's marriage—Comparative quiet
of Italy 295
CHAPTER XIV
State of the papacy at the election of Alexander VI.—His
election, character, and children—The aspect of Italy at the
close of her golden age—The disputed succession of Naples
reopened—Character and views of Charles VIII.—Proposed
league to oppose him frustrated—State of the Roman
Campagna—The old and new military systems in Italy 315
CHAPTER XV
Italy ill prepared for the French invasion—Duke Guidobaldo sent
against the Orsini—Lucrezia Borgia's second marriage—Descent
of Charles VIII.—He reaches Naples and retreats—Battle of the
Taro—The Duke engaged in the Pisan war—Is taken prisoner by
the Orsini and ransomed 341
CHAPTER XVI
The crimes and ambition of the Borgia—Murder of the Duke of
Gandia—Duke Guidobaldo's expeditions against Perugia and
Tuscany—He adopts Francesco Maria della Rovere as his heir—
Louis XII. succeeds to Charles VIII., and to his views upon Italy
—Cesare Borgia created Duke Valentino—Duke Guidobaldo at
Venice 363
CHAPTER XVII
The condition of Romagna—Cesare Borgia overruns and seizes
upon it—The spirit of his government—Naples invaded by Louis,
and handed over to Spain—Lucrezia Borgia's fourth marriage 379
CHAPTER XVIII
Duke Guidobaldo's retired life—Cesare Borgia surprises and
seizes Urbino—The Duke's flight—The diet of La Magione—
Rising in the Duchy, and his return—He again retires 399

APPENDICES

I.Poetry of the family of Montefeltro 427

II. Inventory of articles taken by Brigida Sueva di
Montefeltro, alias Sister Serafina, into the Convent of
Corpus Domini 433
III.Poetry of Ottaviano degli Ubaldini 436
IV. Instrument containing the concessions demanded by
the citizens and acceded to by Count Federigo, on
being chosen as their Seigneur 438
V. Devices and mottoes of the Dukes of Urbino 443
VI. The illuminated MSS. in the Urbino Library 446
VII. Duke Federigo of Urbino a Knight of the Garter 450
VIII. The army of Charles VIII., in 1493 460
IX. The battle of the Taro, in 1495 463
X. The arrival of Duke Valentino at the French Court 468
XI. Ludovico Sforza's entry into Lyons, in 1500 470
 XII. Sonnet to Italy by Marcello Filosseno 472
XIII. Marriage festivities of Lucrezia Borgia at Ferrara, in
1502 473
Genealogical Tables At end of
book
 ILLUSTRATIONS
James Dennistoun of Dennistoun. From a medallion in
the possession of his nephew James W. Dennistoun of
Dennistoun Frontispiece
TO FACE PAGE
View of Urbino. (Photo Alinari) 22
The Battle of S. Egidio. After the picture by Paolo
Uccello in the National Gallery. Portraits of Carlo
Malatesta and his nephew Galeotto "il Beato" 44
Leonello d'Este. After the picture by Pisanello in the
Morelli Gallery, Bergamo. (Photo Alinari) 54
Nicolò Piccinino. From a bronze medal by Pisanello. By
the courtesy of G.F. Hill, Esq. 70
Vittorino da Feltre. From a medal by Pisanello in the
British Museum. By the courtesy of G.F. Hill, Esq. 70
San Leo and Maiuolo. From a drawing by Agostino Nini 78
Federigo of Urbino. From the XV. Century relief in the
Bargello, Florence. (Photo Alinari) 86
Francesco Sforza. From the XV. Century relief in the
Bargello, Florence. (Photo Alinari) 98
Federigo, Duke of Urbino, and Battista, his wife. From
the picture by Piero della Francesca in the Uffizi Gallery,
Florence. (Photo Alinari) 120
Allegory. After the picture by Piero della Francesca in
the Uffizi Gallery, Florence. (Photo Alinari) 122
Allegory. After the picture by Piero della Francesca in
the Uffizi Gallery, Florence. (Photo Alinari) 124
Sigismondo Malatesta. Detail from the fresco by Piero
della Francesca in the Tempio Malatestiano in Rimini.
(Photo Alinari) 132
Urbino. From an original drawing by Agostino Nini of
Bologna 148
The Flagellation. After the picture by Piero della
Francesca in the Sacristy of the Duomo, Urbino.
Supposed portraits of Duke Federigo and Caterino
Zeno. (Photo Alinari) 152
Fifteenth-century Court of the Palazzo Ducale, Urbino.
(Photo Alinari) 162
Pio II. at Ancona. After the fresco by Pinturicchio in the
Cathedral Library, Siena. (Photo Brogi) 178
Portrait of Leon Battista Alberti. From the relief by
Pisanello in the Dreyfus Collection 194
Pope Sixtus IV. From a miniature prefixed to the
dedication copy of Platina's Lives of the Popes in the
Vatican Library 202
Battista Sforza, Duchess of Urbino, second wife of Duke
Federigo. From the bust by Francesco Laurana in the
Bargello, Florence. (Photo Alinari) 214
Federigo of Urbino and his Family. Detail from the
picture by Justus of Ghent, in the Palazzo Ducale,
Urbino. (From the Ducal Collection.) (Photo Alinari) 216
Lorenzo de' Medici. From the fresco by Ghirlandaio in S.
Trinità, Florence. (Photo Alinari) 238
Giuliano de' Medici. (Photo Alinari) 240
The Birth of Venus. Supposed portrait of Simonetta
Cattaneo—mistress of Giuliano de' Medici. Detail from
the picture by Sandro Botticelli in the Uffizi Gallery,
Florence. (Photo Alinari) 242
Astorgio III. de' Manfredi. From the picture by Scaletti
in the Pinacoteca of Faenza 258
Federigo di Montefeltro. After the picture by Justus of
Ghent, once in the Ducal Collection at Urbino, now in
the Palazzo Barberini in Rome. (Photo Anderson) 266
The Contessa Palma of Urbino. After the portrait by
Piero della Francesca in the National Gallery 280
Guidobaldo I. From a picture in the Colonna Gallery in
Rome 296
Caterina Sforza. After the picture by Marco Palmezzani
in the Pinacoteca of Forlì. (Photo Alinari) 306
Isabella of Aragon. After the drawing by Beltraffio in
the Biblioteca Ambrogiana, Milan. (Photo Anderson) 310
Pope Alexander VI. Detail from a fresco by Pinturicchio
in the Borgia apartments of the Vatican, Rome 320
"Diva Julia." From a bronze medal ca. 1482 by L'Antico
in the British Museum. By the courtesy of G.F. Hill, Esq. 330
Cesare Borgia. From a medal ca. 1500 in the British
Museum. By the courtesy of G.F. Hill, Esq. 330
Julius II as Cardinal. From a medal in the British
Museum. By the courtesy of G.F. Hill, Esq. 330
St. Catherine of Alexandria. Supposed portrait of
Lucrezia Borgia by Pinturicchio. Detail from a fresco in
the Borgia apartments of the Vatican, Rome. (Photo
Anderson) 344
Bianca, daughter of Ludovico Sforza. After the picture
by Ambrogio de' Predis in the Biblioteca Ambrogiana,
Milan. (Photo Anderson) 352
Cesare Borgia as the Emperor. Detail from the fresco of
the Disputa of S. Catherine in the Borgia apartments of
the Vatican. (Photo Anderson) 364
 CHRONOLOGICAL TABLE
CHAPTER I

A.D. PAGE
The duchy of Urbino, how composed 3
Its characteristic features, and traditional
topography 4
Origin of Italian communities 4
Rise of the Guelph and Ghibelline factions 5
Counts of the empire 6
Republics established in Italy 7
Opinions regarding their spirit 8
The seigneurs attain to sovereignty 10
Practical distinction of Guelph and Ghibelline 11
Early military system 12
Origin and influence of free companies 14
The term Republic misapplied 15
Their principle of liberty examined 16
Political divisions of Romagna and La Marca in
the fifteenth century 18
Opposed to modern speculations and the aims
of Young Italy 19
Mariotti's admissions regarding freedom 20

CHAPTER II

Examples of these ideas in the dynasties of Urbino 22

1160. The early Counts of Montefeltro are invested with
Urbino 22
1371. Invited to Cagli 22
1384. Received at Gubbio 22
1433. Acquired Casteldurante 23
1445. Purchased Fossombrone 23
1474. Sinigaglia given to the della Rovere 23
1513. They obtained Pesaro and Gradara 23
Statistics of the state so composed 23
1160- Its dynastic changes
1631. 24
Early genealogy of the Montefeltri 24
1160- The Counts of Carpegna
1815. 25
1154. Antonio, first Count of Montefeltro 25
1216. Buonconte, first Count of Urbino 25
1268. Count Guido the Elder, his prowess 26
1282. Takes Forlì by stratagem 27
1289. Excommunicated as a Ghibelline 27
1296. Abdicates and becomes a friar 28
1294. Abdication of Celestine V. 28
” Succeeded by Boniface VIII. 28
1296. His feuds with the Colonna 29
” He recalls Count Guido to the world 30
” Dante's confession of the Count 30
” How far consistent with fact 32
” The Count's piety attested by Boniface 33
1298. Sept. His death at Assisi
27. 34
1300. The struggles of his successors 35
1377. Antonio Count of Urbino 36
1384. Extends his sway over Gubbio, Cagli, and Cantiano 37
1390. His mild government and literary tastes 37
1404. May 9. His death announced to the authorities of Siena
by his son 38
” His children 39
” His daughter Battista, wife of Galeazzo Malatesta,
Lord of Pesaro 39
” Her literary acquirements 40
” Battista takes the veil 40
” Misfortunes of her daughter Elisabetta 41

CHAPTER III

1404. Guidantonio Count of Urbino 42

1408. Made Lord of Assisi 42
1413. And Vice-general of Romagna 43
” Braccio di Montone 43
1417. Nov. Election of Pope Martin V.
11. 44
1418. Dec. Count Guidantonio made Duke of Spoleto 44
1420. Braccio reconciled to the Pope 45
1424. March The Count marries Caterina Colonna
4. 45
” His disputes with the Brancaleoni 45
1430. Sept. Made Captain-general of Florence
3. 46
1431. March Election of Pope Eugenius IV.
3. 46
1438. Oct. 9. Death of Countess Caterina 47
1442. Feb. Death of Count Guidantonio
20. 47
” His children 47
” His daughter Brigida Sueva's singular history 48
” His natural children 49
” Origin of the Ubaldini della Carda 49
 ” Notice of Ottaviano Ubaldino 50
1424. Birth of Count Oddantonio of Urbino 51
1443. April Made Duke of Urbino
26. 51
” His vicious career 52
1444. July His assassination
22. 53
” His intended marriage 55
1439- Two original letters from him to the magistrates of
1443. Siena 56
The dukedom lapsed on his death 58

CHAPTER IV

Federigo Count of Urbino 61

1422. June 7. The mystery and misstatements regarding his
birth 61
1424. Dec. Set at rest by his legitimation
22. 62
” The Brancaleoni of Mercatello 63
1430. Their heiress Gentile betrothed to Count Federigo 64
” The state of Italy at this time 64
” Rome and the Papacy 65
” Florence and Central Italy 66
” Lombardy and Venice 67
1433. Federigo sent to Venice as a hostage 68
1434. Made a companion of the Hose 68
” Becomes a pupil of Vittorino da Feltre at Mantua 69
” Character and system of Vittorino 70
1433. Federigo knighted by the Emperor 71
1437. Dec. 2. His marriage 72
” Nicolò Piccinino successor of Braccio di Montone 72
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

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.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookname.com