An Introduction to Computational Science Allen

Holder pdf download

https://textbookfull.com/product/an-introduction-to-
computational-science-allen-holder/

Download more ebook from https://textbookfull.com

We believe these products will be a great fit for you. Click
the link to download now, or visit textbookfull.com
to discover even more!

An Introduction to the Crusades S.J. Allen

https://textbookfull.com/product/an-introduction-to-the-crusades-
s-j-allen/

Computational Materials Science: An Introduction,

Second Edition Lee

https://textbookfull.com/product/computational-materials-science-
an-introduction-second-edition-lee/

An Introduction to Computational Origami Tetsuo Ida

https://textbookfull.com/product/an-introduction-to-
computational-origami-tetsuo-ida/

An Introduction to Computational Macroeconomics

Economic Methodology Anelí Bongers

https://textbookfull.com/product/an-introduction-to-
computational-macroeconomics-economic-methodology-aneli-bongers/
Introduction to Interdisciplinary Studies Allen F.
Repko

https://textbookfull.com/product/introduction-to-
interdisciplinary-studies-allen-f-repko/

An introduction to electrical science Second Edition

Waygood

https://textbookfull.com/product/an-introduction-to-electrical-
science-second-edition-waygood/

CS 101 An Introduction to Computational Thinking 1st

Edition Sarbo Roy

https://textbookfull.com/product/cs-101-an-introduction-to-
computational-thinking-1st-edition-sarbo-roy/

An Introduction to Physical Science 15th Edition James

Shipman

https://textbookfull.com/product/an-introduction-to-physical-
science-15th-edition-james-shipman/

Python Programming An Introduction to Computer Science

John M. Zelle

https://textbookfull.com/product/python-programming-an-
introduction-to-computer-science-john-m-zelle/
International Series in
Operations Research & Management Science

Allen Holder
Joseph Eichholz

An Introduction to
Computational
Science
International Series in Operations Research
& Management Science

Volume 278

Series Editor
Camille C. Price
Department of Computer Science, Stephen F. Austin State University,
Nacogdoches, TX, USA

Associate Editor
Joe Zhu
Foisie Business School, Worcester Polytechnic Institute, Worcester, MA, USA

Founding Editor
Frederick S. Hillier
Stanford University, Stanford, CA, USA
More information about this series at http://www.springer.com/series/6161
Allen Holder • Joseph Eichholz

An Introduction to
Computational Science

123
Allen Holder Joseph Eichholz
Department of Mathematics Department of Mathematics
Rose-Hulman Institute of Technology Rose-Hulman Institute of Technology
Terre Haute, IN, USA Terre Haute, IN, USA

Additional material to this book can be downloaded from http://extras.springer.com.

ISSN 0884-8289 ISSN 2214-7934 (electronic)

International Series in Operations Research & Management Science
ISBN 978-3-030-15677-0 ISBN 978-3-030-15679-4 (eBook)
https://doi.org/10.1007/978-3-030-15679-4

c Springer Nature Switzerland AG 2019

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

This Springer imprint is published by the registered company Springer Nature Switzerland AG.
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Dedicated to the educators who have so
nurtured our lives.

Joseph Eichholz: Weimin Han

Allen Holder: Harvey J. Greenberg
and
Jeffrey L. Stuart.
Foreword

The interplay among mathematics, computation, and application in science

and engineering, including modern machine learning (formerly referred to as
artificial intelligence), is growing stronger all the time. As such, an educa-
tion in computational science has become extremely important for students
pursing educations in STEM-related fields. This book, along with undergrad-
uate courses based on it, provides a modern and cogent introduction to this
rapidly evolving subject.
Students pursuing undergraduate educations in STEM-related fields are
exposed to calculus, differential equations, and linear algebra. They are also
exposed to coursework in computer science in which they learn about com-
plexity and some algorithmic theory. These courses commonly require stu-
dents to implement algorithms in popular languages such as C, Java, Python,
MATLAB, R, or Julia. However, many students experience unfortunate dis-
connects between what is learned in math classes and what is learned in
computer science classes. This book, which is designed to support one or two
undergraduate terms of study, very nicely fills this gap.
The book introduces a wide spectrum of important topics in mathemat-
ics, spanning data science, statistics, optimization, differential equations, and
randomness. These topics are presented in a precise, dare I say rigorous,
mathematical fashion. However, what makes this book special is that it in-
tertwines mathematical precision and detail with an education in how to
implement algorithms on a computer. The default mathematical language is
MATLAB/Octave, but the interface to other programming languages is also
discussed in some detail.
The book is filled with interesting examples that illustrate and motivate
the ideas being developed. Let me mention a few that I found particularly
interesting. One homework exercise asks students to access “local” tempera-
ture data from the National Oceanic and Atmospheric Association’s website
and to use that data as input to a regression model to see if average daily

vii
viii Foreword

temperatures are on the rise at that location. Another exercise in the chapter
on modern parallel computing requires students to generate the famous frac-
tal called the Mandelbrot set. Yet another exercise in the material on linear
regression models asks students to reformulate a nonlinear regression model
that computes an estimate of the center and radius of a set of random data
that lies around a circle. The reformulation is linear, which allows students
to solve the problem with their own code from earlier exercises.
The computational part of the book assumes the reader has access to either
MATLAB or Octave (or both). MATLAB is perhaps the most widely used
scripting language for the type of scientific computing addressed in the book.
Octave is freely downloadable and is very similar to MATLAB. The book
includes an extensive collection of exercises, most of which involve writing
code in either of these two languages. The exercises are very interesting, and
a reader/student who invests the effort to solve a significant fraction of these
exercises will become well educated in computational science.

Robert J. Vanderbei
Preface

The intent of this text is to provide an introduction to the growing interdisci-

plinary field of Computational Science. The impacts of computational stud-
ies throughout science and engineering are immense, and the future points
to an ever increasing reliance on computational ability. However, while the
demand for computational expertise is high, the specific topics and curricula
that comprise the arena of computational science are not well defined. This
lack of clarity is partially due to the fact that different scientific problems
lend themselves to different numerical methods and to diverse mathematical
models, making it difficult to narrow the broad variety of computational ma-
terial. Our aim is to introduce the “go to” models and numerical methods
that would typically be attempted before advancing to more sophisticated
constructs if necessary. Our goal is not to pivot into a deep mathematical
discussion on any particular topic but is instead to motivate a working savvy
that practically applies to numerous problems. This means we want students
to practice with numerical methods and models so that they know what
to expect. Our governing pedagogy is that students should understand the
rudiments of answers to: How expensive is a calculation, how trustworthy is a
calculation, and how might we model a problem to apply a desired numerical
method?
We mathematically justify the results of the text, but we do so without
undue rigor, which is appropriate in an undergraduate introduction. This
often means we assert mathematical facts and then interpret them. The au-
thors share the melancholy of a matter-of-fact presentation with many of our
mathematical colleagues, but we hope that students will continue to pursue
advanced coursework to complete the mosaic of mathematical justification.
Our humble approach here is to frame the future educational discourse in a
way that provides a well-honed skill for those who might consider working in
computational science.

ix
x Preface

The intended audience is the undergraduate who has completed her or his
introductory coursework in mathematics and computer science. Our general
aim is a student who has completed calculus along with either a traditional
course in ordinary differential equations or linear algebra. An introductory
course in a modern computer language is also assumed. We introduce topics
so that any student with a firm command of calculus and programming should
be able to approach the material. Most of our numerical work is completed
in MATLAB, or its free counterpart Octave, as these computational plat-
forms are standard in engineering and science. Other computational resources
are introduced to broaden awareness. We also introduce parallel computing,
which can then be used to supplement other concepts.
The text is written in two parts. The first introduces essential numerical
methods and canvases computational elements from calculus, linear algebra,
differential equations, statistics, and optimization. Part I is designed to pro-
vide a succinct, one-term inauguration into the primary routines on which
a further study of computational science rests. The material is organized so
that the transition to computational science from coursework in calculus, dif-
ferential equations, and linear algebra is natural. Beyond the mathematical
and computational content of Part I, students will gain proficiency with el-
emental programming constructs and visualization, which are presented in
their MATLAB syntax. Part II addresses modeling, and the goal is to have
students build computational models, compute solutions, and report their
findings. The models purposely intersect numerous areas of science and engi-
neering to demonstrate the pervasive role played by computational science.
This part is also written to fill a one-term course that builds from the compu-
tational background of Part I. While the authors teach most of the material
over two (10 week) terms, we have attempted to modularize the presentation
to facilitate single-term courses that might combine elements from both parts.

Terre Haute, IN, USA Allen Holder

Terre Haute, IN, USA Joseph Eichholz
Summer 2018
Acknowledgments

Nearly all of this text has been vetted by Rose-Hulman students, and these
careful readings have identified several areas of improvement and pinpointed
numerous typographical errors. The authors appreciate everyone who has
toiled through earlier drafts, and we hope that you have gained as much
from us as we have gained from you.
The authors have received aid from several colleagues, including many who
have suffered long conversations as we have honed content. We thank Eivind
Almaas, Mike DeVasher, Fred Haan, Leanne Holder, Jeff Leader, John Mc-
Sweeney, Omid Nohadani, Adam Nolte, and Eric Reyes. Of these, Jeff Leader,
Eric Reyes, and John McSweeney deserve special note. Dr. Leader initiated
the curriculum that motivated this text, and he proofread several of our ear-
lier versions. Dr. Reyes’ statistical assistance has been invaluable, and he
was surely tired of our statistical discourse. Dr. McSweeney counseled our
stochastic presentation. Allen Holder also thanks Matthias Ehrgott for sab-
batical support. Both authors are thankful for Robert Vanderbei’s authorship
of the Foreword.
The professional effort to author this text has been a proverbial labor of
love, and while the authors have enjoyed the activity, the task has impacted
our broader lives. We especially want to acknowledge the continued support
of our families—Amanda Kozak, Leanne Holder, Ridge Holder, and Rowyn
Holder. Please know that your love and support are at the forefront of our
appreciation.
We lastly thank Camille Price, Neil Levine, and the team at Springer.
Their persistent nudging and encouragement has been much appreciated.

xi
Contents

Part I Computational Methods

1 Solving Single Variable Equations . . . . . . . . . . . . . . . . . . . . . . . . 3

1.1 Bisection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 Linear Interpolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3 The Method of Secants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4 Newton’s Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.4.1 Improving Efficiency with Polynomials . . . . . . . . . . . . . . 15
1.4.2 Convergence of Newton’s Method . . . . . . . . . . . . . . . . . . 17
1.4.3 MATLAB Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

2 Solving Systems of Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

2.1 Systems of Linear Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.1.1 Upper- and Lower-Triangular Linear Systems . . . . . . . . 34
2.1.2 General m × n Linear Systems . . . . . . . . . . . . . . . . . . . . . 36
2.2 Special Structure: Positive Definite Systems . . . . . . . . . . . . . . . . 44
2.2.1 Cholesky Factorization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.2.2 The Method of Conjugate Directions . . . . . . . . . . . . . . . . 49
2.3 Newton’s Method for Systems of Equations . . . . . . . . . . . . . . . . 52
2.4 MATLAB Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

3 Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.1 Linear Models and the Method of Least Squares . . . . . . . . . . . . 67
3.1.1 Lagrange Polynomials: An Exact Fit . . . . . . . . . . . . . . . . 72
3.2 Linear Regression: A Statistical Perspective . . . . . . . . . . . . . . . . 74
3.2.1 Random Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.2.2 Stochastic Analysis and Regression . . . . . . . . . . . . . . . . . 83

xiii
xiv Contents

3.3 Cubic Splines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

3.4 Principal Component Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
3.4.1 Principal Component Analysis and the Singular
Value Decomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
3.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

4 Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
4.1 Unconstrained Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
4.1.1 The Search Direction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
4.1.2 The Line Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
4.1.3 Example Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
4.2 Constrained Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
4.2.1 Linear and Quadratic Programming . . . . . . . . . . . . . . . . 151
4.3 Global Optimization and Heuristics . . . . . . . . . . . . . . . . . . . . . . . 164
4.3.1 Simulated Annealing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
4.3.2 Genetic Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
4.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177

5 Ordinary Differential Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

5.1 Euler Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
5.2 Runge-Kutta Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
5.3 Quantifying Error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
5.4 Stiff Ordinary Differential Equations . . . . . . . . . . . . . . . . . . . . . . 207
5.5 Adaptive Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
5.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219

6 Stochastic Methods and Simulation . . . . . . . . . . . . . . . . . . . . . . . 237

6.1 Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
6.2 Numerical Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
6.2.1 Simpson’s Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
6.2.2 Monte Carlo Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
6.2.3 Bootstrapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
6.2.4 Deterministic or Stochastic Approximation . . . . . . . . . . 251
6.3 Random Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
6.3.1 Simulation and Stochastic Differential Equations . . . . . 255
6.3.2 Simulation and Stochastic Optimization Models . . . . . . 259
6.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262

7 Computing Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269

7.1 Language Choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
7.2 C/C++ Extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
7.3 Parallel Computing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
7.3.1 Taking Advantage of Built-In Commands . . . . . . . . . . . . 277
7.3.2 Parallel Computing in MATLAB and Python . . . . . . . . 278
7.3.3 Parallel Computing in Python . . . . . . . . . . . . . . . . . . . . . 284
Contents xv

7.3.4 Pipelining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287

7.3.5 Ahmdal’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
7.3.6 GPU Computing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
7.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296

Part II Computational Modeling

8 Modeling with Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309

8.1 Signal Processing and the Discrete Fourier Transform . . . . . . . 310
8.1.1 Linear Time Invariant Filters . . . . . . . . . . . . . . . . . . . . . . 311
8.1.2 The Discrete Fourier Transform . . . . . . . . . . . . . . . . . . . . 315
8.1.3 The Fast Fourier Transform . . . . . . . . . . . . . . . . . . . . . . . 320
8.1.4 Filtering Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
8.1.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
8.2 Radiotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
8.2.1 A Radiobiological Model to Calculate Dose . . . . . . . . . . 326
8.2.2 Treatment Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
8.2.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
8.3 Aeronautic Lift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
8.3.1 Air Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
8.3.2 Flow Around a Wing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
8.3.3 Numerical Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
8.3.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353

9 Modeling with Ordinary Differential Equations . . . . . . . . . . . 355

9.1 Couette Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
9.1.1 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
9.2 Pharmacokinetics: Insulin Injections . . . . . . . . . . . . . . . . . . . . . . 363
9.2.1 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
9.3 Chemical Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
9.3.1 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375

10 Modeling with Delay Differential Equations . . . . . . . . . . . . . . . 377

10.1 Is a Delay Model Necessary or Appropriate? . . . . . . . . . . . . . . . 378
10.2 Epidemiology Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
10.3 The El-Niño–La-Niña Oscillation . . . . . . . . . . . . . . . . . . . . . . . . . 383
10.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385

11 Partial Differential Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389

11.1 The Heat Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
11.2 Explicit Solutions by Finite Differences . . . . . . . . . . . . . . . . . . . . 392
11.3 The Wave Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
11.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
xvi Contents

12 Modeling with Optimization and Simulation . . . . . . . . . . . . . . 403

12.1 Stock Pricing and Portfolio Selection . . . . . . . . . . . . . . . . . . . . . . 403
12.1.1 Stock Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404
12.1.2 Portfolio Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
12.1.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
12.2 Magnetic Phase Transitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
12.2.1 The Gibbs Distribution of Statistical Mechanics . . . . . . 419
12.2.2 Simulation and the Ising Model . . . . . . . . . . . . . . . . . . . . 421
12.2.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428

13 Regression Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431

13.1 Stepwise Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
13.2 Qualitative Inputs and Indicator Variables . . . . . . . . . . . . . . . . . 437
13.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441

A Matrix Algebra and Calculus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447

A.1 Matrix Algebra Motivated with Polynomial Approximation . . 447
A.2 Properties of Matrix–Matrix Multiplication . . . . . . . . . . . . . . . . 453
A.3 Solving Systems, Eigenvalues, and Differential Equations . . . . 455
A.3.1 The Nature of Solutions to Linear Systems . . . . . . . . . . 456
A.3.2 Eigenvalues and Eigenvectors . . . . . . . . . . . . . . . . . . . . . . 460
A.4 Some Additional Calculus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
 Part I
Computational Methods
Chapter 1
Solving Single Variable
Equations

No more fiction for us: we calculate; but that we may calculate, we had to make
fiction first. – Friedrich Nietzsche

The problem of solving the equation f (x) = 0 is among the most storied
in all of mathematics, and it is with this problem that we initiate our study
of computational science. We assume functions√have their natural domains in
the real numbers. For instance, a function like x exists over the collection of
nonnegative reals. The right-hand side being zero is not generally restrictive
since solving either f (x) = k or g(x) = h(x) can be re-expressed as f (x)−k =
0 or g(x) − h(x) = 0. Hence looking for roots provides a general method to
solve equations.
The equation f (x) = 0 may have a variety of solutions or none at all.
For example, the equation x2 + 1 = 0 has no solution over the reals, the
equation x3 + 1 = 0 has a single solution, and the equation sin(x) = 0 has
an infinite number of solutions. Such differences complicate searching for a
root, or multiple roots, especially without forehand knowledge of f . Most
algorithms attempt to find a single solution, and this is the case we consider.
If more than one solution is desired, then the algorithm can be re-applied in
an attempt to locate others.
Four different algorithms are presented, those being the time-tested meth-
ods of bisection, secants, interpolation, and Newton. Each has advantages
and disadvantages, and collectively they are broadly applicable. At least one
typically suffices for the computational need at hand. Students are encour-
aged to study and implement fail-safe programs for each, as they are often
useful.

© Springer Nature Switzerland AG 2019 3

A. Holder, J. Eichholz, An Introduction to Computational Science,
International Series in Operations Research & Management Science 278,
https://doi.org/10.1007/978-3-030-15679-4 1
4 1 Solving Single Variable Equations

1.1 Bisection

The theoretical underpinnings of bisection lie in one of the most important

theorems of Calculus, that being the Intermediate Value Theorem. This the-
orem ensures that a continuous function attains all intermediate values over
an interval.
Theorem 1 (Intermediate Value Theorem). Let f (x) be continuous on
the interval [a, b], and let k be between f (a) and f (b). Then, there is a c
between a and b such that f (c) = k.
The method of bisection is premised on the assumption that f is continuous
and that f (a) and f (b) differ in sign. We make the tacit assumption that
a < b, from which the Intermediate Value Theorem then guarantees a solution
to f (x) = 0 over the interval [a, b]. To illustrate, the function f (x) = x3 − x
satisfies
f (1.5) = 1.875 > 0 and f (−2) = −6 < 0.
So f is guaranteed to have at least one root over the interval [−2, 1.5]. Indeed,
f has three roots over this interval, although the Intermediate Value Theorem
only guarantees one.
Bisection proceeds by halving the interval. The midpoint of [−2, 1.5] is
−0.25, and since f (−0.25) = 0.2344 > 0, we again have from the Interme-
diate Value Theorem that f has a root over the interval [−2, −0.25]. Two
observations at this point are (1) this new interval removes two of the three
roots, and (2) we are guaranteed to be within (2 − (−1.5))/2 = 3.5/2 of a
solution to f (x) = 0. The algorithm continues by halving the new interval
and removing the half that no longer guarantees a root. Bisection is called a
bracketing method since a solution is always “bracketed” between two other
values. The first five iterations for this example are listed in Table 1.1. The
iterates show that the interval collapses onto the solution x = −1. These
iterations are depicted in Fig. 1.1, and general pseudocode is listed in Algo-
rithm 1.
One of the benefits of bisection is that it has a guarantee of how close
it is to a root. The algorithm’s estimate of the root for the example above
is within (1.5 + 2)/25 = 0.1094 of the real solution after the fifth iteration.
Indeed, after k iterations the algorithm is within (1.5 + 2)/2k of the root, a
fact that allows us to calculate how many iterations are needed to ensure a
desired accuracy. So if we want to be within 10−8 of a solution, i.e. 8 digits
of accuracy, then finding the first integer value of k such that 3.5/2k ≤ 10−8
shows that bisection will need
 
ln(3.5 × 108 )
k= = 29
ln(2)

iterations for this example.

1.1 Bisection 5

Iteration Interval Best value of |f |

0 [−2.000000, 1.500000] 1.875000
1 [−2.000000, −0.250000] 0.234375
2 [−1.125000, −0.250000] 0.234375
3 [−1.125000, −0.687500] 0.298828
4 [−1.125000, −0.906250] 0.161957
5 [−1.015625, −0.906250] 0.031986
Table 1.1 Five iterations of bisection solving f (x) = x3
− x = 0 over [−2, 1.5]. The best
value of f in the third column is the smallest value of |f (x)| at the end points of the
iteration’s interval.

Fig. 1.1 An illustration of bisection Fig. 1.2 The function fq (x) = x1/(2q+1)
solving f (x) = x3 − x = 0 over [−2, 1.5]. for q = 0, 1, 5, 10, and 50. The red dots
The intervals of each iteration are shown show the value of fq (x) at the approxi-
in red at the bottom of the figure, and mate solution to fq (x) = 0 after 15 itera-
vertical green lines are midpoint evalua- tions of bisection initiated with [0.5, 1.3].
tions.

While the distance to a root can be controlled by the number of iterations,

this fact does not ensure that the function itself is near zero. For example,
the function fq (x) = x1/(2q+1) is increasingly steep at x = 0 as the integer q
increases, see Fig. 1.2. Indeed, for any x between −1 and 1 other than zero
we have  
 
lim x1/(2q+1)  = 1.
q→∞

So any approximate solution to fq (x) = 0 can have a function value near

1 for large q. In other words, x being near a solution does not ensure that
f (x) is near 0. If we start with the interval [−0.5, 1.3], then 15 iterations
guarantee a solution within a tolerance of 10−4 and end with the interval
[−1.2 × 10−5 , 4.3 × 10−5 ] independent of q. However, for q = 1, 5, 10, and
6 1 Solving Single Variable Equations

Algorithm 1 Pseudocode for the method of bisection

k=1
while unmet termination criteria do
x̂ = (a + b)/2
if sign(f (x̂)) == sign(f (a)) then
a = x̂
else if sign(f (x̂)) == sign(f (b)) then
b = x̂
else if sign(f (x̂)) == 0 then
set termination status to true
else
set status to failure and terminate
end if
k =k+1
end while
return best estimate of root

50 the values of fq at the midpoint of this interval are, respectively, 0.0247,

0.3643, 0.5892, and 0.8959, see Fig. 1.2.
Alternatively, the value of f (x) can be close to zero even though x remains
some distance from a solution. For example, functions of the form
x
gq (x) =
1 + (10q x)2

have the property that gq (x) → 0 as q → ∞ for all nonzero x. So any

real number can appear to be a solution for sufficiently large q. Suppose the
termination criterion is |gq (x)| ≤ 10−4 . If bisection is initialized with the
interval [−0.9, 0.8], then the terminal interval is [−1.95 × 10−4 , 1.2 × 10−5 ]
for q = 2. The value of gq at the midpoint of this interval is reasonable
at 1.2 × 10−5 . However, for q = 3 the value at the initial left end point
is gq (−0.9) = −1 × 10−6 , and bisection terminates immediately with an
estimated solution of x = −0.9, which is some distance from the desired
x = 0.
A dilemma of writing code for general use is deciding the termination
criteria. Two common convergence conditions are

|f (x)| ≤ ε, for the best computed value of x, and

|x − x∗ | ≤ δ, where x∗ is a solution to f (x) = 0,

where ε and δ are user supplied convergence tolerances. Satisfying both is

preferred to satisfying one, but requiring both could be computationally bur-
densome and possibly unrealistic.
1.2 Linear Interpolation 7

Care is warranted when implementing pseudocode like that of Algorithm 1.

The termination criteria should include tests for the original end points, for
the value of the function at the newly computed midpoint, and for the width
of the interval. Moreover, we might add tests to
• ensure that the original interval is well defined, i.e. a < b,
• see that the signs of f (a) and f (b) differ, and
• validate that the function value is real, which is not always obvious in
MATLAB and Octave, e.g. (-1)^(1/3) isn’t real, but nthroot(-1,3) is.
Returning a status variable in addition to the computed solution lets a user
know if she or he should trust the outcome. In many cases a status variable
encodes one of several termination statuses. Moreover, it is generally good
advice to undertake the prudent step of verifying the assumed properties of
the input data and of the computed result, even though doing so lengthens
code development and can seem to pollute a clean, succinct, and logical pro-
gression. However, including measures to vet computational entities will lead
to trustworthy algorithms that can be used for diverse tasks.

1.2 Linear Interpolation

The method of linear interpolation is similar to that of bisection, but in this

case the algorithm uses the function’s values at the endpoints and not just
their signs. As with bisection, the method of linear interpolation uses the
Intermediate Value Theorem to guarantee the existence of a root within an
interval. Unlike the method of bisection, the method of linear interpolation
approximates the function to estimate the next iterate. If the function is itself
linear, or more generally affine, then linear interpolation will converge in one
step, an outcome not guaranteed with bisection.
Assume that f is continuous on the interval [a, b] and that f (a) and f (b)
differ in sign. The line through the points (a, f (a)) and (b, f (b)) is called the
interpolant and is

f (a) − f (b)
y − f (a) = (x − a).
a−b
Solving for x with y = 0 gives the solution x̂ so that
 
a−b
x̂ = a − f (a) . (1.1)
f (a) − f (b)

If f (x̂) agrees in sign with f (a), then the interval [x̂, b] contains a root and
the process repeats by replacing a with x̂. If f (x̂) agrees in sign with f (b),
then the interval [a, x̂] contains a root and the process repeats by replacing
8 1 Solving Single Variable Equations

b with x̂. If |f (x̂)| is sufficiently small, then the process terminates since x̂ is
a computed root. The pseudocode for linear interpolation is in Algorithm 2.

Algorithm 2 Pseudocode for the method of linear interpolation

k=1
while unmet termination criteria do
x̂ = a − f (a)(a − b)/(f (a) − f (b))
if sign(f (x̂)) == 0 then
set termination status to true
else if sign(f (x̂)) == sign(f (a)) then
a = x̂
else if sign(f (x̂)) == sign(f (b)) then
b = x̂
else
set status to failure and terminate
end if
k =k+1
end while
return best estimate of root

The first five iterations of linear interpolation solving f (x) = x3 − x = 0

over the interval [−2, 1.5] are listed in Table 1.2, and the first few iterations
are illustrated in Fig. 1.3. Linear interpolation converges to x = 1 as it solves
f (x) = x3 − x = 0 initiated with [−2, 1.5], whereas the method of bisection
converges to x = −1. So while the methods are algorithmically similar, the
change in the update produces different solutions. The convergence is “one-
sided” in this example because the right end point remains unchanged. Such
convergence is not atypical, and it can lead to poor performance. As a second
example, iterations solving f (x) = cos(x) − x = 0 over [−0.5, 4.0] are listed
in Table 1.2 and are shown in Fig. 1.4. In this case convergence is fairly rapid
due to the suitable linear approximation of f by the interpolant.
Interpolation does not provide a bound on the width of the k-th iterate’s
interval. In fact, there is no guarantee that the width of the intervals will
converge to zero as the algorithm proceeds. Therefore, the only reasonable
convergence criterion is |f (x)| < ε. As with bisection, care and prudence are
encouraged during implementation.

1.3 The Method of Secants

The method of secants uses the same linear approximation as the method
of linear interpolation, but it removes the mathematical certainty of captur-
ing a root within an interval. One advantage is that the function need not
1.3 The Method of Secants 9

Fig. 1.3 An illustration of the first few Fig. 1.4 An illustration of the first few
iterations of linear interpolation solving iterations of linear interpolation solving
f (x) = x3 − x = 0 over [−2, 1.5]. The red f (x) = cos(x) − x = 0 over [−0.5, 4]. The
lines at the bottom show how the inter- red lines at the bottom show how the
vals update according to the roots of the intervals update according to the roots
linear interpolants (shown in magenta). of the linear interpolants (shown in ma-
genta).

f (x) = x3 − x f (x) = cos(x) − x

Iteration Interval Best |f | Interval Best |f |
0 [−2.0000, 1.5000] 1.8750 [−0.5000, 4.0000] 1.3776
1 [ 0.6667, 1.5000] 0.3704 [ 0.5278, 4.0000] 0.0336
2 [ 0.8041, 1.5000] 0.2842 [ 0.5278, 0.7617] 0.0336
3 [ 0.8957, 1.5000] 0.1771 [ 0.7379, 0.7617] 0.0019
4 [ 0.9479, 1.5000] 0.0963 [ 0.7391, 0.7617] 0.0000
5 [ 0.9748, 1.5000] 0.0484 [ 0.7391, 0.7617] 0.0000
Table 1.2 The first five iterations of linear interpolation solving x3 − x = 0 over [−2, 1.5]
and cos(x) − x = 0 over [−0.5, 4.0].

change sign over the original interval. A disadvantage is that the new iterate
might not provide an improvement without the certainty guaranteed by the
Intermediate Value Theorem.
The method of secants is a transition from the method of linear interpo-
lation toward Newton’s method, which is developed in Sect. 1.4. The mathe-
matics of the method of secants moves us theoretically from the Intermediate
Value Theorem to the Mean Value Theorem.
Theorem 2 (Mean Value Theorem). Assume f is continuous on [a, b]
and differentiable on (a, b). Then, for any x in [a, b] there is c in (a, x) such
that
f (x) = f (a) + f  (c)(x − a).
10 1 Solving Single Variable Equations

This statement of the Mean Value Theorem is different than what is typi-
cally offered in calculus, but note that if x = b, then we have the common
observation that for some c in (a, b),

f (b) − f (a)
f  (c) = .
b−a
The statement in Theorem 2 highlights that the Mean Value Theorem is a
direct application of Taylor’s Theorem, a result discussed more completely
later.
The method of secants approximates f  (c) with the ratio

f (b) − f (a)
f  (c) ≈ ,
b−a
which suggests that

f (b) − f (a)
f (x) ≈ f (a) + (x − a) .
b−a
The method of secants iteratively replaces the equation f (x) = 0 with the
approximate equation

f (b) − f (a)
0 = f (a) + (x − a) ,
b−a
which gives a solution of
 
b−a
x̂ = a − f (a) .
f (b) − f (a)

This update is identical to (1.1), and hence the iteration to calculate the new
potential root is the same as the method of linear interpolation. What changes
is that we no longer check the sign of f at the updated value. Pseudocode
for the method of secants is in Algorithm 3.
The iterates from the method of secants can stray, and unlike the methods
of bisection and interpolation, there is no guarantee of a diminishing interval
as the algorithm progresses. The value of f (x) can thus worsen as the algo-
rithm continues, and for this reason it is sensible to track the best calculated
solution. That is, we should track the iterate xbest that has the nearest func-
tion value to zero among those calculated. Sometimes migrating outside the
original interval is innocuous or even beneficial. Consider, for instance, the
first three iterations of solving f (x) = x3 − x = 0 over the interval [−2, 1.5]
in Table 1.3 for the methods of secants and linear interpolation. The algo-
rithms are the same as long as the interval of linear interpolation agrees with
the last two iterates of secants, which is the case for the first two updates.
The third update of secants in iteration 2 is not contained in the interval
[a, b] = [0.6667, 0.8041] because the function no longer changes sign at the
1.3 The Method of Secants 11

Algorithm 3 Pseudocode for the method of secants

k=1
while unmet termination criteria do
if f (b) == f (a) then
set status to failure and terminate
else
x̂ = a − f (a)(a − b)/(f (a) − f (b))
a=b
b = x̂
k =k+1
end if
end while
return best estimate of root

Linear interpolation Secants

Iteration Interval Update (x̂) a b Update (x̂)
0 [−2.0000, 1.5000] 0.6667 −2.000 1.5000 0.6667
1 [ 0.6667, 1.5000] 0.8041 1.5000 0.6667 0.8041
2 [ 0.8041, 1.5000] 0.8957 0.6667 0.8041 1.2572
3 [ 0.8957, 1.5000] 0.9479 0.8041 1.2572 0.9311

Table 1.3 The first three iterations of linear interpolation and the method of secants
solving f (x) = x3 − x = 0 over [−2, 1.5].

end points, and in this case, the algorithm strays outside the interval to es-
timates the root as 1.2572. Notice that if f (a) and f (b) had been closer in
value, then the line through (a, f (a)) and (b, f (b)) would have been flatter.
The resulting update would have been a long way from the earlier iterates.
However, the algorithm for this example instead converges to the root x = 1,
so no harm is done. The first ten iterations are in Table 1.4, and the first few
iterations are depicted in Fig. 1.5.
Widely varying iterates are fairly common with the method of secants,
especially during the initial iterations. If the secants of the most recent iter-
ations favorably point to a solution, then the algorithm will likely converge.
An example is solving f (x) = cos(x)−x = 0 initiated with a = 10 and b = 20,
for which linear interpolation would have been impossible because f (a) and
f (b) agree in sign. The first two iterates provide improvements, but the third
does not. Even so, the algorithm continues and converges with |f (x)| < 10−4
in seven iterations, see Table 1.5. Indeed, the method converges quickly once
the algorithm sufficiently approximates its terminal solution. This behavior is
12 1 Solving Single Variable Equations

Iteration a b x̂ x best f (x best )

0 −2.0000 1.5000 0.6667 0.6667 −0.3704
1 1.5000 0.6667 0.8041 0.8041 −0.2842
2 0.6667 0.8041 1.2572 0.8041 −0.2842
3 0.8041 1.2572 0.9311 0.9311 −0.1239
4 1.2572 0.9311 0.9784 0.9784 −0.0418
5 0.9311 0.9784 1.0025 1.0025 0.0050
6 0.9784 1.0025 0.9999 0.9999 −0.0002
7 1.0025 0.9999 1.0000 1.0000 −0.0000
Table 1.4 Iterations of the method of secants solving f (x) = x3 − x = 0 initiated with
a = −2.0 and b = 1.5.

Fig. 1.5 The first three iterations of the secant method solving f (x) = x3 −x = 0 initiated
with a = −2.0 and b = 1.5. The third secant line produces an iterate outside the interval
of the previous two.

typical and concomitant with the favorable convergence of Newton’s method,

which is being mimicked (see Sect. 1.4.2).
The splaying of unfavorable iterates does not always lead to convergence,
and in some cases the algorithm diverges. An example is solving f (x) =
x2 /(1 + x2 ) = 0 initiated with a = −2 and b = 1. In this case the magnitude
of the iterates continues to increase, and our best solution remains x = 1
with a function value of f (1) = 0.5. Iterations are listed in Table 1.5.
The method of secants often converges if the initial points are sufficiently
close to the root. However, what constitutes close enough varies from function
to function and is usually difficult to determine, particularly without prior
knowledge of the function and its solution. It is typical to employ a “try
and see” strategy with the method of secants. We start with two reasonable
guesses of the root and simply see if the method converges.
1.4 Newton’s Method 13

Solving cos(x) − x = 0 Solving x2 /(1 + x2 ) = 0

Initiated with a = 10 and b = 20 Initiated with a = −2 and b = 1
Iteration Iterate (x̂) xbest f (xbest ) Iterate (x̂) xbest f (xbest )
1 −2.3835 −2.3835 1.6574 6.0000 1.0 0.5
2 −0.6377 −0.6377 1.4412 −4.2857 1.0 0.5
3 11.0002 −0.6377 1.4412 −400.71 1.0 0.5
4 0.7109 0.7109 0.0469 7277.9 1.0 0.5
5 0.7546 0.7546 −0.0260 −1.2 × 109 1.0 0.5
6 0.7390 0.7390 0.0002 6.6 × 1016 1.0 0.5
7 0.7391 0.7391 0.0000

Table 1.5 The third iterate of the method of secants solving cos(x) − x = 0 lacks im-
provement, but the algorithm continues to converge. The iterates attempting to solve
x2 /(1 + x2 ) = 0 diverge.

1.4 Newton’s Method

Newton’s method is a numerical stalwart and is commonly the method of

choice. Unlike the method of secants, which estimates f  , Newton’s method
uses the derivative itself to estimate the function locally. Assuming f is con-
tinuously differentiable, we have from the Mean Value Theorem (Theorem 2
in the previous section) that for any x there is a c between x and a such that

f (x) = f (a) + f  (c)(x − a).

So, if x is close to a, then f is approximated by the linear function

f (x) ≈ f (a) + f  (a)(x − a).

The substitution of f  (a) for f  (c) points to the analytic assumption that f 
be continuous.
Newton’s method iteratively replaces the equation f (x) = 0 with the ap-
proximate equation
f (a) + f  (a)(x − a) = 0,
which has the solution
f (a)
x̂ = a −
f  (a)
as long as f  (a) = 0. The iterate x̂ is the new estimate of the root, and unless
the process reaches a termination criterion, x̂ replaces a and the process
continues. Pseudocode for Newton’s method is in Algorithm 4. The first six
iterates solving x3 − x = 0 and cos(x) − x = 0 are listed in Table 1.6. Several
of the first iterations are illustrated in Figs. 1.6 and 1.7.
14 1 Solving Single Variable Equations

Algorithm 4 Pseudocode for Newton’s method

while unmet termination criteria do
if f  (a) = 0 then
a = a − f (a)/f  (a)
else
set status to failure and terminate
end if
end while

Solving x3 − x = 0 Solving cos(x) − x = 0

Initiated with a = −0.48 Initiated with a = −1
Iteration Iterate (a) f (a) Iterate (a) f (a)
0 −0.4800 0.3694 1.0000 1.5403
1 0.7163 −0.3488 8.7162 −9.4755
2 1.3632 1.1702 2.9761 −3.9624
3 1.1075 0.2508 −0.4258 1.3365
4 1.0139 0.0283 1.8512 −2.1279
5 1.0003 0.0006 0.7660 −0.0454
6 1.0000 0.0000 0.7392 −0.0002

Table 1.6 The first six iterates of Newton’s method solving x3 − x = 0 initialized with
x = −0.48 and cos(x) − x = 0 initiated with x = −1.

Fig. 1.6 An illustration of the first three Fig. 1.7 An illustration of the first
iterates of Newton’s method solving x3 − three iterates of Newton’s method solving
x = 0 initiated with x = −0.48. The cos(x) − x = 0 initiated with x = −1.0.
tangent approximations are the magenta The tangent approximations are the ma-
lines. genta lines.
1.4 Newton’s Method 15

The calculation of x̂ assumes that f  (a) is not zero, and indeed, this is
a computational concern. If r is a root and f  (r) = 0, then the assumed
continuity of f  ensures that |f  (a)| is arbitrarily small as the process con-
verges to r. Division by small values challenges numerical computation since
it can result in less accurate values or exceed the maximum possible value
on the computing platform. A reasonable assumption for both computational
and theoretical developments is that f  (r) = 0, and a reasonable termination
criterion is to halt the process if |f  (a)| is sufficiently small.
A notational convenience is to let Δx = x̂ − a, which succinctly expresses
the update as
x̂ = a + Δx, where f  (a) Δx = −f (a).
The change between iterations, Δx, is called the Newton step. The expression
f  (a)Δx = −f (a) removes the algebraic concern of dividing by zero, and it
nicely extends to the higher dimensional problems of the next chapter. Of
course, the solution is unique if and only if f  (a) = 0.
Newton’s method shares a loss of guaranteed convergence with the method
of secants, and iterates can similarly stray during the search for a solution.
However, it is true that if the starting point is close enough to the root, then
Newton’s method will converge. A common approach is to “hope for the best”
when using Newton’s method, and for this reason the exit statuses of both
the method of secants and Newton’s method are particularly critical.

1.4.1 Improving Efficiency with Polynomials

The most computationally expensive calculation of solving f (x) = 0 with an

iterative method is commonly the evaluation of the function itself. A func-
tion’s evaluation in many real-world problems depends on substantial calcu-
lations that can take hours, days, weeks, or months. This reality suggests that
we pay particular attention to the cases for which the computational burden
can be reduced, and we demonstrate how to hasten the evaluation of poly-
nomials. Restricting our attention to polynomials may seem quaint against
the backdrop of possible functions that we might consider, but polynomials
are regularly used to approximate more complicated functions. Hence, an
efficient calculation method finds wide application.
Evaluating polynomials and their derivatives can be accomplished more
efficiently than is commonly first perceived. We consider the quadratic f (x) =
3x2 + 2x + 1 as a small example. Evaluating f (x) as expressed requires 3
multiplications and 2 additions: one multiplication for x2 , another for 3x2 ,
and one for 2x. However, we may rewrite the polynomial as

f (x) = 3x2 + 2x + 1 = (3x + 2)x + 1, (1.2)

Discovering Diverse Content Through
Random Scribd Documents
Doge. The steady pursuance of commercial aggrandisement brought Venice
in the previous reign into conflict with Bologna. Duties were levied on all
ships trading in the ports between Ravenna and Fiume, and a captain of the
Gulf of Adria appointed to exact them by force of arms. It was only too
apparent that Venice aimed at making the whole Adriatic a Venetian sea, and
in 1283 the territories of the Republic in Dalmatia were menaced by a
formidable coalition of Aquileia, Ancona, and the Count of Goritz.
The Republic replied by laying siege to Trieste. During the armistice
granted by Morosini, the Venetian commander, to enable the Patriarch of
Aquileia to bury his nephew, fallen in battle, a certain Contestabile of
Infantry, Gerard of the Long Lances, was found to have corresponded with
the enemy by means of slips of paper attached to his arrows. He confessed
under torture and was shot from a mangonel into the enemy’s camp, a
mangled mass of treachery. The siege was, however, a failure, and Morosini
on his return was disgraced and imprisoned. The victorious allies marched
on Caorle, captured the podestà and burned his palace: they even singed the
very mane of the Lion of St Mark by a descent on Malamocco. The
executive, endowed with unlimited powers, made a levy en masse on the
whole able-bodied population, another armament was fitted out and at length
Trieste fell. The Pope, anxious to buttress the tottering fabric of the Latin
dominion in the East, effected a peace, and a huge bonfire in the Piazza
made of the surrendered Triestine artillery satisfied Venetian pride, though
her attempt to dominate the Adriatic was but partially successful.
The sleepless eyes of Venetian statesmen were now turned southwards.
Constance of Swabia, unhappy Manfred’s

“bella figlia, genitrice

dell’ onor di Cicilia e d’ Aragona,”[26]

by her marriage brought the strong arm of Peter of Aragon to enforce her
claims to the throne of the Two Sicilies, which the Pope’s darling, Charles of
Anjou, had won by the defeat of Manfred and the destruction of the
Ghibelline cause at Benevento.
The Republic, still smarting under the commercial condominium of the
Genoese at Constantinople, and enticed by the prospect of regaining her
former ascendency, was drawn into an offensive and defensive alliance with
the Pope and Charles against the Greeks. But on Easter Eve, in 1282, an
insult offered to a noble Sicilian damsel by a French soldier fired the rage of
the Sicilians, and before the dawn of Easter Day the blood of eight thousand
French in Palermo alone had glutted their vengeance. The Sicilian Vespers
wrecked Charles’s fortunes, and the Republic turned to renew her former
understanding with the Greek Emperor. Martin IV., enraged at the defection
of his ally, laid Venice under the ban of the Church, but his opportune death
soon made reconciliation possible.

DOGE’S PALACE—THE JUDGMENT OF SOLOMON CORNER

Visitors to Venice may have noticed a traditional motto written in brass

nails on some of the gondolas that ply for hire which admirably sums up the
old popular Venetian idea of government: Pane in Piazza: Giustizia in
Palazzo (Bread in the market-place: Justice in the palace). This principle
always inspired the rule of the aristocracy. In 1284, during Giovanni
Dandolo’s reign, a calamitous inundation had plunged the people into
misery. To meet their urgent needs a loan was raised on Government
security, and ten thousand bushels of wheat were distributed at a nominal
cost among the people.
In October of the same year the mint issued the first gold ducat of Venice,
which for centuries was famous for its purity, fineness and weight
throughout the whole commercial world. Orders were given that it was to be
similar to and even purer than the golden florin of Florence, which had been
coined thirty-two years before. Sanudo remembers to have seen an
inscription on marble in the mint, dated 1285, commemorating the first
striking of the gold ducat of the Venetians in honour of the Blessed Mark the
Evangelist and of all saints, in the reign of the renowned Doge of the
Venetians, Giovanni Dandolo. This beautiful zecchino (sequin) was worth
about nine shillings and sixpence in English money, and admirably
illustrates the dress of the Doges during a period of five hundred and thirteen
years. The evolution of the corno or horn on the ducal bonnet may be clearly
traced on the coins issued between the reigns of Francesco Foscari and
Leonardo Loredano.
The sepulchre had hardly received the body of Giovanni Dandolo in 1289
when a formidable demonstration in favour of Giacomo, son of Doge
Lorenzo Tiepolo, put the new constitution to a severe test. Giacomo, son and
grandson of Doges, was known to have popular leanings, and so threatening
was the attitude of the crowd on the Piazza that the Privy Council personally
urged him to disclaim any intention of accepting the proffered honour.
Tiepolo, preferring his country’s peace to the gratification of his own
ambition, exhorted the crowd to respect the law, and left for the mainland till
the crisis should be past. But the delicate electoral machinery was never for a
moment put out of gear. The provisional Government was appointed;
through all the tumult the electors calmly rattled their ballots in the Ducal
Palace, and to the sullen displeasure of the popular party a prominent
aristocrat, Pietro Gradenigo, was proclaimed Doge twenty-three days after
the death of his predecessor. For the first time the officer who recited the
formula: Quest’è il vostro doge si vi piacerà, turned aside without staying to
receive the approbation of the people.
The long reign of Gradenigo (1289-1311) is one of the most important in
the annals of Venice. By the fall of Acre in 1291 the doom of the Christian
power in the Holy Land was sealed, and Venice, whose interest in the
Crusades and in the Latin dominion over Syria was frankly a commercial
one, turned the new situation to her own advantage. Her policy was to
frustrate her rivals, the Genoese. To the scandal of Christendom a treaty was
concluded with the infidel in 1299, and ere long slaves and materials of war
were openly sold by the Venetians in their ports. The Sultan declared in the
charter his steadfast will that the Venetians should be protected and
honoured beyond all people in the world, and entitled to the sole right of a
Saracen escort for Christian pilgrims to Jerusalem. The Genoese found
themselves squeezed out of the coast towns, and Venice in exclusive
possession of the Syrian trade.

ON THE GRAND CANAL.

The Pope, anxiously revolving the sad vicissitudes of the Christians in the
east, turned to Venice and Genoa, praying them for the love of Christ to
combine and save the fair island of Cyprus, still unpolluted by the presence
of the infidels. But the lion of St Mark was a fierce yoke-fellow. The more
restricted the field of influence became between Venice and Genoa the more
bitter grew their jealousy. Two fleets were, however, fitted out in response to
the Papal appeal. Their prows had scarcely touched Cyprian waters when a
fight took place between some of the allied ships, and to the edification of
the Saracen the two greatest maritime powers of Christendom were soon
engaged in mutual destruction. Unavailing efforts were made by the Church
to heal the strife, for while the Dominican envoys were treating at Venice the
feverish activity at the arsenal told too plainly that the time for the
peacemaker was not yet come. Rumours soon reached Venice of an alliance
between the Genoese and the Greeks and of the threatened closure of the
Dardanelles to her ships. She delayed no longer to strike. All her seamen
between sixteen and sixty were enrolled; her patrician houses were called to
furnish their part of a new armament, and on October 7th, 1294, the fleet
was under sail. The admiral, Marco Besegio, sighted the Genoese fleet under
Nicolo Spinola off Ayas, in Asia Minor. The enemy was inferior in strength,
and Besegio, too confident perhaps of victory, was out-manœuvred, defeated
with heavy losses, and himself slain. The Genoese, to clinch their victory,
despatched a mighty fleet of nigh two hundred sail manned by forty-five
thousand men, among whom were the chiefs of their noble houses.
Meanwhile Venice, shrewdly calculating that the heavy financial strain
involved in the maintenance of so huge an armament would soon wear the
enemy out, steadily equipped a new fleet, called out a fresh levy, and
concentrated her force on the defence of the lagoons. Before a year was past
the Genoese, after spoiling and slaughtering the hapless inhabitants of Canea
in Crete, returned to port.
Early in 1295 news came to Venice which stung her to fury. A street row
at Constantinople had developed into a general attack by the Genoese on the
Venetians. The former were victorious, and after flinging the Venetian
Governor out of the windows of the palace, dashing him to pieces,
proceeded to an indiscriminate massacre of the inhabitants of the Venetian
colony. The Greeks sent envoys to disclaim any responsibility in the outrage,
but they were hectored by the Doge, who demanded an enormous indemnity,
which served but to cement their alliance with the Genoese. Late in the
spring the Venetian commander, Ruggieri Morosini, with a fleet of forty
galleys, forced the Dardanelles, wasted the Genoese suburb of Galata and
laid siege to Constantinople. Meanwhile another fleet, under Giov. Soranzo,
entered the Black Sea and sacked the Genoese settlement of Caffa; but the
elements amply avenged the Genoese. Soranzo returned to Venice bearing an
unheroic story of vessels disabled and men frozen to death by the rigours of
an Euxine winter. The year 1297 passed in petty expeditions, and towards
the end of the autumn Boniface VIII. essayed to negotiate a peace. The
magnanimous Pope (Dante’s pet enemy) went so far as to offer, if the
Genoese paid one-half, to pay himself the other half of the claims of the
Venetians, but the latter rejected all compromise, and Boniface despairing of
success inculpated the pride of Venice, and washed his hands of the whole
business. Each power prepared for a final struggle.
Among the wealthy Venetians whose enthusiasm took the form of
offering themselves and their ships to the common cause was a certain
Marco Polo but recently returned from adventurous journies in the
mysterious lands of the Grand Khan of Tartary; in Persia, China, Japan, and
the Indies; and who from his wonderful stories of the million peopled cities
and millions of jewels and treasure he had seen in his twenty-five years’
wanderings was popularly known as Messer Marco Milione. In August 1298
all was ready and a fleet of ninety-five sail, under the command of Andrea
Dandolo, set its course southwards and came upon the Genoese squadron of
eighty-five vessels, under Lamba Doria off the island of Curzola. The fleets
were about evenly matched, and on September 8th the action began. Doria,
by superior seamanship, got the weather gauge and the Venetians, fighting
too with the sun in their eyes, were routed. Twelve galleys, whose captains,
panic-stricken, had abandoned the fight, alone escaped. With abject mien
they told the extent of the disaster. The fine fleet was sunken, captured or
burned, the loss in killed appalling, and seven thousand of their countrymen
were on their way to Genoese prisons.[27] Among the captives was Messer
Marco Milione himself, who to relieve the tedium of his imprisonment,
dictated in halting French to his prison comrade Rustichello the story of his
wanderings and adventures. A small court, in which Marco Polo’s house
stood on a site now covered by the Malibran Theatre, is called to this day the
Corte del Milione.
Venice, conscious that her staying power was greater than her rival’s,
without a moment of panic set about equipping another fleet of a hundred
galleys. But Genoa, exhausted by her costly victory, was willing to treat, and
in 1299 the Imperial Vicar of Milan effected a peace between the Republics
on honourable terms.
During the Genoese war the aristocracy had quietly matured plans for
fencing off their preserves from any intrusion of the democracy. Two
abortive attempts had been made in 1286 and 1296 to restrict membership of
the Great Council to members of the aristocracy. Gradenigo, who was a
leader of iron will and indomitable purpose, succeeded the next year in
achieving the revolution in the Constitution known as the shutting of the
Great Council (Serrata del Gran Consiglio). The Quarantia[28] were charged
(1) to put to the ballot one by one all who for the
past four years had sat in the Great Council.
Those who received not less than twelve votes
were to be members up to Michaelmas, when,
after being subjected to a new ballot, they were
to serve for a further period of one year. (2)
Three electors were to be appointed who should
submit further names of non-members of the
Great Council for election under the same
system of ballot. (3) The three were to sit in the
Council until Michaelmas, when they were to be
superseded by other three, who should sit for a
year. (4) The law could not be repealed save
with the consent of five of the six privy
councillors, twenty-five of the Quarantia, and REMAINS OF MARCO
two-thirds of the Great Council. Such were the POLO’S HOUSE
chief provisions of the measure which
transformed the aristocracy into an oligarchy and created the Maggior
Consiglio.
 DOGE’S PALACE—SALA DEL MAGGIOR CONSIGLIO

It will be seen that by the second clause an avenue was left open by
which a Venetian, not a member of the favoured class, might enter the
aristocratic close, but it was rendered inoperative by the principle which the
three laid down for their guidance, that only those whose paternal ancestors
had been members of the Great Council between 1172 and 1297 should be
eligible for ballot. The effect of the change was to increase the number of the
Council. In 1296 it consisted of two hundred and ten members. In 1311 they
had risen to ten hundred and seventeen; in 1340 to twelve hundred and
twelve; in 1490 to fifteen hundred and seventy; and in 1510 to sixteen
hundred and seventy-one.
In 1315 it was enacted that a book be kept for the inscription of the names
of all persons above eighteen years of age who had the right to enter the
Council. So keen was the ambition to be inscribed that in the year following
a fine of thirty lire was imposed on all those whose names were unlawfully
entered and who did not remove them within a month. In 1319 Avvogadori, a
sort of heraldic officers, were appointed and charged to subject to the
severest scrutiny the titles of applicants for inscription, and in order to
frustrate any attempt to tamper with the electors it was ordained that as many
ballots should be used as there were names inscribed, and that of these, a
number equal to the candidates to be elected should be of gold. The names
were read out in the order of their entry and a boy extracted a ballot as each
man was called. Those to whom the gold ballots fell were declared elected.
It was further ordained that after the lapse of two years all who had reached
the age of twenty-five years and were in possession of the necessary
qualifications should ipso facto be entitled to enter the Council. Thus the
electors’ functions ended, and any descendant of an aristocratic family who
fulfilled the conditions required by the law became at that age a member of
the Great Council. This was the actual and definite Serrata (Nov. 25th,
1319).
Every noble was bound to notify his marriage and the birth of his children
at the Avvogaria to be entered in a book and stringent regulations were from
time to time laid down to insure the purity of the family record. Owing to the
association of the golden ballots with the right to enter the Council, this book
was called the Libro d’Oro, the Golden book. The Council elected all
officers of State, imposed taxes, decreed laws, made peace and war,
concluded alliances, until owing to its unwieldy growth many of its powers
were delegated to the Senate, the Council retaining as its chief function the
election of the officers of the Republic. The Senate was definitely
established in 1230 and consisted of sixty members, nominated by four
electors of the Great Council, to whom later other sixty were added called
the Zonta or addition. The Consiglio Minore (Privy Council) was still
composed of six members chosen from the wards of the city. They, with the
Doge, presided over the Senate, and with the three chiefs of the Quarantia
formed the Serenissima Signoria (Signory). They could act in the absence of
the Doge but the Doge could take no action without them. They opened
dispatches, received petitions, prepared the agenda for the Great Council and
the Senate, read the Coronation oath every year to the Doge and if need were
admonished him. The Quarantia (Council of forty) was the judicial authority
and controlled the Mint, heard complaints from the subject cities and
provinces, and gave audience to ambassadors. In the fifteenth century the
civil and criminal functions of the Quarantia were separated and two
Quarantie established. The Doge, the living embodiment of the Republic,
presided over all these assemblies. In 1308 a small giunta of seven Savii
(wise men or experts) was formed to deal with Ferrarese affairs and did its
work so well that it continued in office. It was subsequently enlarged and
subdivided in 1442 into three bodies, one dealing with home affairs, one
with mainland affairs, the third with the arsenal. The three united formed the
Collegio (Cabinet). By the permanent appointment of the Council of Ten in
1335 was evolved the famous constitution of Venice which for its stability
and efficiency became the admiration of every statesman in Europe, and
filled with envy the Italian states of the mainland.

BOATS AT ANCHOR.

CHAPTER VII

The Oligarchy—Commercial supremacy—The Bajamonte Conspiracy—The Council of the Ten—The

Prisons

“O thou that art situate at the entry of the sea which art a merchant of the people for
many isles, ... thou hast said: I am of perfect beauty. Thy borders are in the midst of the seas,
thy builders have perfected thy beauty.... Thy wise men were thy pilots.... All the ships of
the sea were in thee to occupy thy merchandise.... Syria was thy merchant ... they occupied
in thy fairs with emeralds, purple and broidered work, fine linen and coral and agate.”—
Ezekiel.

THE fourteenth century opens the era of the oligarchy. Venice had made
peace with the only rival that could challenge her maritime supremacy. She
had not yet entangled herself in an aggressive continental policy. The tramp
of the advancing Turk was too far away to echo in the lagoons. The wealth
of the Indies and of the far East flowed through her markets. Her merchants
laid the known world under contribution. “By way of the Syrian ports and of
Alexandria came the cloves, nutmegs, mace and ebony of the Moluccas; the
sandal wood of Timor; the costly camphor of Borneo; the benzoin of
Sumatra and Java; the aloes, wood of Cochin China; the perfumes, gums,
spices, silks and innumerable curiosities of China, Japan and Siam; the
rubies of Pegu; the fine fabrics of Coromandel; the richer stuffs of Bengal;
the spikenard of Nepaul and Bhutan; the diamonds of Golconda; the
Damascus steel of Nirmul; the pearls, sapphires, topazes and cinnamon of
Ceylon; the pepper, ginger and satin wood of Malabar; the lac, agates and
sumptuous brocades and jewelry of Cambay; the costus and graven vessels,
wrought arms and broidered shawls of Cashmere; the bdellium of Scinde;
the musk of Thibet; the galbanum of Khorossan; the assafœtida of
Afghanistan; the sagapenum of Persia; the ambergris, civet and ivory from
Zanzibar; the myrrh, balsam and frankincense of Zeila, Berbera and
Shehr.”[29] The bare recital of this catalogue has the effect of a poem and
fills the imagination with visions of Oriental splendour. Every year six
trading fleets averaging about five hundred vessels each sailed, one for the
Black Sea, another for Greece and Constantinople; others for the Syrian
ports; for Egypt, Barbary and North Africa; for Flanders and England. These
ships were the property of the State, and in due time a public crier
announced the number of galleasses ready for the annual voyages. They
were farmed out to the highest bidders, who were required to prove their
qualifications and the amount of their capital, and to provide on each
galleasse accommodation, a suitable mess, and space for a small cargo, for
eight young nobles, who thus were trained in naval science and gained
experience of commerce. The vessels were constructed on fixed models and
convertible at will into men-of-war. Every man aboard, passenger or seaman,
bore arms, and was compelled to fight the ship in case of attack.
Standardised fittings were obtainable at every Venetian maritime station to
replace any that might be lost or damaged by storm or battle. The food and
comfort of the seamen were carefully provided for. A cross painted or carved
on the side served as a load-line, and Government inspectors checked any
attempt to overload. Each ship carried a band of music.
The cargo of a Syrian or Egyptian galleasse was worth about two hundred
thousand[30] ducats, and it has been estimated that the Republic in the
fifteenth century could dispose of three thousand three hundred ships, thirty-
six thousand seamen and sixteen thousand shipwrights. The consuls at every
Venetian port were charged to inspect the weights and measures of the
traders and to prevent adulteration or fraud. If the consul were found to be
venal he was branded on the forehead. At home the same measures were
taken to maintain the standard of quality. In 1550 English woollen goods
from the Thames were exposed with the brand of the Senate upon them in St
Mark’s Palace as evidence of English dishonesty and the decay of English
faith. In the fifteenth century, when Turkish pirates infested the seas, navi
armate (war-ships) were built to convoy the merchant fleets, and a state
navy was thus formed in which slaves or criminals were forced to work the
oars.
In 1556, sixty out of a gang of Lutherans convicted of heresy, marching
through Flanders on their way to the Venetian galleys, were rescued from
slavery by the people of Maestricht and their guards stoned. In earlier times,
before the navy was differentiated into merchant and war-ships, Sclavonians
were employed for this exhausting labour, and those who manned the
galleasses to Flanders and England possessed a burial vault in North
Stoneham Church near Southampton.
In the event of a naval war a levy was made on all the male inhabitants.
Those liable were divided into groups of twelve and lots were drawn to
decide who should serve first. The unfit provided substitutes or were fined.
Those on service were given free bread rations and were paid five lire a
month by the Commune and one lira from each man of the twelve who was
not chosen. Romanin estimates that the equipment of a galley in the early
thirteenth century was equal to that of a frigate of seventy-four guns in his
day (1850). The discipline was perfect. The seamen were said to obey their
chiefs as sheep do their shepherd. Gambling or swearing was severely
punished by flogging.
“When the hour of departure neared, the Commander came on board
preceded by trumpeters and followed by his staff. Perfect silence reigned as
he began his inspection. Every man was at his post, every oar in its place. All
the arms, accoutrements and appointments were carefully examined, and
when the trumpeters gave signal for departure the rowers simultaneously
plied their oars, or if the wind were fair threw them up and sails[31] were set
with marvellous alertness. A general holiday was observed with great pomp
and magnificence. The ships were coloured white and vermilion, the sails
bright with variegated stripes, the poop was richly gilt, the figure-head of the
most beautiful design. The Doge and his Council with dazzling pageantry,
senators in scarlet robes, the élite of Venetian ladies, famed for grace and
beauty and arrayed in gorgeous dresses, and hundreds of citizens in gay
gondolas witnessed the departure.”[32] A stirring scene, surpassed alone by
that when the clanging of bells from St Mark’s tower called the people to
view a victorious fleet sailing up the lagoons and the enemy’s standards
trailing on the waters.
In Gradenigo’s reign we note the first indication of a policy of territorial
aggrandisement on the Italian mainland. Little wars had already been waged:
with the Paduans in 1142, to prevent the diversion of the course of the
Brenta; and with Ferrara in 1240, to maintain trading privileges. In 1308
these were again endangered by a dispute between rival claimants for the
lordship of Ferrara. Venice intervened and was brought into conflict with the
Pope. His Holiness, as a temporal enemy, fought at a vantage, for to the
material bolts of Mars he was able to add the spiritual thunders of the
Church. When the Papal warning was received, the Doge addressed the
Councillors, and stoutly defended his policy and told them that they were not
children to be frightened by words. There was an angry scene in the council
chamber, and for once the ominous cries, “Guelph and Ghibelline,” were
heard within the walls of the palace. The ducal party maintained their
position and the ban was laid on Venice. The Doge, his Councillors and the
citizens were excommunicated, their possessions in Ferrara confiscated,
every treaty with them declared void, commercial relations forbidden, and
all the clergy summoned to leave. The Pope’s words, however, were winged
with terror to the Venetians. News soon came of banks, factories and ships
sacked in Italy, France and England, and even in far Asia. Their trade, except
with the infidel, was paralysed; religious and civic life disintegrated. But the
Republic never winced. On the very day that the papal interdict reached
Venice instructions were sent to the Venetian podestà at Ferrara to fortify
himself in Castel Tebaldo and manfully and potently to uphold the rights and
honour of his country.
The Venetian garrison, however, weakened by disease, surrendered after a
long struggle, and met the fate of the vanquished. The fleet was destroyed,
and growing unrest in the State forced the Doge and his party to make terms
with the Pope. Ferrara was acknowledged to be a papal fief, and an
indemnity paid for the restoration of the trading privileges of the Republic.
In the year of the Serrata the corpses of Bocconi, a popular leader, and
ten of his followers dangled between the red columns as a warning to the
disaffected, but after the inglorious issue of the war the discontent of the
people was intensified, and found a rallying-point in certain ambitious and
disgraced nobles of the Quirini, Tiepolo and Badoer[33] families, who were
united by a common hatred of the ducal party. Secret meetings were held in
Casa Quirini near the Rialto, and Bajamonte, the people’s darling, the “Gran
Cavaliere,” son of Jacopo Tiepolo, was drawn into the conspiracy. It was
determined to organise a revolution and assassinate the Doge and his chief
supporters. The insurrection was fixed for Sunday, June 14th, 1310, the eve
of St Vito’s Day. Down the two main avenues of traffic that debouch from
the north on the Piazza, the Calle dei Fabri and the Merceria, two divisions
under the leadership of Marco Quirini, the chief conspirator, and of
Bajamonte, were to march and simultaneously attack the palace, meanwhile
Badoer was sent to collect sympathisers at Padua. All had been foreseen save
the treachery of man and of the elements. In the early dawn, as the
revolutionists rushed from Casa Quirini, shouting “Liberty” and “Death to
Doge Gradenigo,” their faces were lashed by a driving rain, their voices
smothered by peals of thunder and the howling of the wind. The movements
failed to synchronise, and the Quirini section encountered in the Piazza, not
their allies from the Merceria, but a ducal force which scattered them and
slew their leader and his son. Men who will betray the State will betray their
fellows. The plot had been divulged by one Marco Donato, and the Doge
had met the danger with his wonted courage and alertness. He increased his
guards, summoned help from Chioggia, Murano and Torcello, called out the
arsenal men, armed his Councillors and their servants. Having disposed of
the Quirini, the Doge was able to deal with Bajamonte’s division in the
Merceria. During the fighting Bajamonte’s standard-bearer met the fate of
Abimelech.[34] A woman aimed a stone mortar from an upper window at
him: it struck him on the head, and the bearer and the banner inscribed with
the word, “Liberty,” fell to the ground. Panic seized the rebels and they fled
across the Rialto bridge. Meanwhile the remnant of Quirini’s party rallied
and made a stand on the Campo S. Lucia, only to be finally crushed by
members of the Painters’ Guild and of the Guild of Charity. A more serious
task remained, to subdue Bajamonte and his followers, who had hewn down
the bridge and fortified themselves in some houses yon side the Rialto. After
many negotiations the rebels surrendered. Their lives were spared, but they
agreed to banish themselves from Venetian territory. Ill-hap, too, had fallen
on Badoer’s reinforcements, which were defeated by the Chioggians. Badoer
and his chief followers were captured and hanged between the red columns.
To perpetuate the memory of this narrow escape from a great peril S. Vito’s
Day was made a day of public festival and thanksgiving for evermore. To
Marco Donato and his descendants was granted membership of the Great
Council. The woman, Lucia Rosso, who had cast the fateful mortar, being
asked to name her reward, begged permission to fly the standard of St Mark
from her window on every feast day, and desired that the procurators of St
Mark, to whom the house belonged, would not raise to her or to her
successors the annual rental of fifteen ducats. The house, known as the Casa
e bottega della grazia del morter, appears from an old painting in the Correr
Museum to have stood on the site of the first house on the left-hand side of
the Merceria entering from the Piazza. The mortar was cast from the third
floor window. “The banner I have seen raised,” says Sanudo, “but now that
the new buildings are made it can no longer be seen from the Piazza. The
under part of Marco Quirini’s house in Rialto was made into shambles, and
there they remain to this day” (about 1520).[35] Bajamonte’s house in S.
Agostino was razed, the site made over to the commune, and a column set up
in the Campo S. Agostino with an inscription stating that the “land once
Bajamonte’s had been confiscated for his wicked treachery and to inspire
others, with terror.”[36] Certain of the marbles of the house were assigned by
the Republic in 1316 for the restoration of the church of S. Vito. For
eighteen years Bajamonte lived in exile and never ceased to plot his revenge
until he was secretly disposed of by an emissary of the Ten.
 THE CLOCK TOWER AND ENTRANCE TO THE MERCERIA.

The Consiglio de’ Dieci shares with the Comité du Salut publique a
sinister notoriety in history. Let us see how far the earlier and more enduring
body deserves its reputation. The great plot had showed the urgent need of
an executive able to act with rapidity and secrecy. The Council of Ten was
appointed to deal with further developments of the plot, but proved so
admirable and effective an instrument that it was more than once renewed
and finally made permanent in 1335. The Ten were charged “to preserve the
liberty and peace of the subjects of the Republic and protect them from the
abuses of personal power.” They were elected by the Great Council with
careful deliberation among the most reputable of the citizens, and no more
than one member of any family could serve. A member sat for one year, he
was not eligible for re-election, he received no pay, he was obliged to retire
if any of his relations were among the accused; it was a capital offence to
receive a gift of any nature. His term of service ended, the dread decemvir
passed again into private life. The Ten elected from themselves three chiefs
(Capi) who served for one month, during which period they were forbidden
to go about the city, to frequent shops or other public places where the
nobility were wont to gather. Among other duties, on the first day of their
month the Capi were required to send to the Signory a list of the prisoners
detained by order of the Ten with suggestions for any reform or
improvement in the prisons, and to take measures to expedite the trial of the
accused. They were to report to the Council all the arrests made by the
previous Capi and to remind the Council of all cases sub judice in the
preceding month. The Doge and his six Privy Councillors were present at the
sittings, and a legal officer without a vote watched the proceedings to check
any abuse of power. Secret denunciations placed in the Bocche del leone,
especially if unsigned, were subject to most elaborate procedure before they
were acted upon. The accused were usually interrogated in darkness, but if
five-sixths of the tribunal agreed, the interrogation might take place in the
light. They could call witnesses. If the minutes of the trial exceeded a
hundred and fifty sheets they were read a second time on another day, that
the members might refresh their strained powers of attention. The defence
was read entire. If the condemnation, after five ballotings, did not command
more than half the votes of the Council, the accused was set at liberty or the
case was retried. When the condemnation had gained an absolute majority it
was subject to four re-ballots before being made final and irrevocable. The
Ten dealt with—criminal charges against nobles; treachery and conspiracy in
the State; espionage; unnatural crimes; secret information likely to be of
advantage to the Republic; the regulations of the Greater Scuole or Guilds;
the use of secret service money; disobedient State officials; false coiners and
debasers of the precious metals used in jewellery; forests and mines; the
glass industry at Murano; acts of violence on the water; the use of arms;
theatres; masked balls and public morals generally; and, after 1692, the
censorship of the printing press.
The tribunal could inflict pecuniary fines; corporal punishment;
banishment, with power to compass his death if the proscribed one were
found outside bounds; imprisonment for any period, and for life; the galleys;
mutilation; death, secretly or publicly. The death sentence was generally
carried out by decapitation or hanging from the columns of the palace or
between the red columns in the Piazzetta. For the more heinous crimes the
guilty were conducted in infamous guise along the Grand Canal, flogged and
broken upon the wheel. Secret executions were rarely resorted to, and
generally with the object of saving the prestige of the nobility by
withdrawing from the public gaze the disgrace of an honoured name.
In 1539 the ever-present dread of Spanish plots fed by the gold of the
New World led to the permanent establishment of Il Supremo Terribile
Tribunale of the Three State Inquisitors. For among the large body of State
officials and members of councils were many patricians who, impoverished
by the decline of commerce, were peculiarly open to corruption, and the
need was felt of a smaller and more expeditious body than the Ten. Of the
Tre Inquisitori di Stato two were appointed by the Ten, one by the Doge’s
Privy Council. The latter sat in the middle clothed in red and was called the
rosso; the former sat one on either side clothed in black, and were known as
the negri. They served for a year and were eligible for re-election. Service
was compulsory under a fine of 500 ducats. Their powers were delegated to
them, as emergencies demanded, by the Ten, who reserved the right of
revising their judgments, which were also published in the Great Council. If
the Three were not unanimous they must refer the case to the Ten. Carefully
indicted rules guarded against the abuse of secret denunciations, and against
the venality or the errors of spies. Suspects were arrested at night and
examined in secret, torture being used in accordance with the usual legal
procedure of the day. Witnesses were also examined in secret by the
Secretary or a ducal notary. The triumvirs acted with appalling swiftness[37]
and secrecy, and stout of heart was he who did not quail when the officer of
the Three touched him on the shoulder with the usual formula, “Their
Excellencies would like to see you.” During the sixteenth century the Ten
and its Committee grew to be the dominant body in the State, until in 1582
the right of calling the Zonta was abolished, and having no longer the power
of associating with them members of any and every council and of spending
money they reverted to their former position.
The tribunals occasionally abused their powers, committed some crimes,
and made errors. The murder of the Carraras was a national sin; the
execution of Foscarini a grievous blunder. But they were a popular body, and
withstood every attack upon them. They were a bulwark against treachery:
they protected the people from the insolence and arbitrariness of nobles: they
maintained equality, and were stern censors of morals. Their best defence is
the fact that they endured to the fall of the Republic.

PONTE DI PAGLIA

Much undeserved obloquy has been cast upon the Ten even by historians
of repute when treating of the famous prisons under their charge, the so-
called Pozzi and Piombi (wells and leads). Lurid pictures have been drawn
of victims tortured in cells hot as furnaces under the leads, and in dungeons
beneath the canal, where neither light nor warmth ever penetrated, and
where the prisoner saw the instruments of his torture on the wall before him.
But in truth the Pozzi were as little underground as the Piombi were
(immediately) under the leads. The Ducal Palace had been furnished with
prisons from its construction by Angelo Participazio and its restoration by
Seb. Ziani until the new prison over the Bridge of Sighs, and beyond the rio
del Palazzo, was completed in 1606. Except the Torreselle (prisons in the
towers), one of which, at the angle of the palace overlooking the Ponte della
Paglia, was demolished in 1532 by order of the Ten, the old prisons were
situated on the east wing of the palace between the inner court and the rio
del Palazzo, and later extended to the other side towards the Molo on the
south. They were on the ground floor, which was sub-divided into two
storeys of cells. Some of the windows looked on the public courtyard, and at
one period the prisoners could talk with passers-by. They were not known as
Pozzi before the seventeenth century. After the erection of the new jail on the
opposite side of the Rio the so-called Pozzi were used for the more
dangerous prisoners, and on the fall of the Republic in 1797 four only were
found therein, scoundrels who richly deserved their fate. The Republic bore
a unique reputation for its humane treatment of prisoners. Zanotto,[38] from
whose admirable monograph we mainly draw, quotes the testimony of Friar
Felice Fabri, who, visiting Venice in the middle of the fifteenth century, was
struck by the merciful treatment of the prisoners of the Republic. In 1443 the
Great Council appointed an advocate to defend the cause of the poor
detained prisoners, and in 1553 a second advocate was chosen. In common
with the whole of Christian Europe Venice used torture to extract
confessions, but she honourably distinguished herself by appointing a
surgeon to examine the prisoners and to report if they were able to bear the
infliction. In 1564 the Ten ordered an infirmary to be prepared for sick
prisoners. The disinfection and cleanliness of the cells and the quality and
quantity of the food and wine supplied to the incarcerated were carefully
inspected.
In 1591 the Senate permitted the Ten to make use of a floor above the
Sala de’ Capi in order that the detained might be in more comfortable, lighter
and better-ventilated cells than those allotted to the condemned. The rooms
were known as the Piombi, since they were on the floor next below the roof,
which was covered with lead. According to Zanotto, between the ceilings,
which were made of a double layer of larch planks, and the roof, was a space
of several yards, varying with the slope of the roof. The rooms were small;
roughly, about twelve by fourteen feet, and from six feet to eight feet high,
and were wainscoted with larch. They were lighted from a corridor, and
ventilators were fixed in the doors. The detained dressed as they pleased,
were allowed to see visitors, and to walk in the corridor.
 S. GIORGIO AND S. MARIA DELLA SALUTE.

Gradenigo died before the papal ban had been removed, and found quiet
sepulture at Murano. A distinguished senator, Stefano Giustiniani, was
chosen his successor, but renounced the office and retired to the monastery
at S. Giorgio. The annalists relate that while the electors sat anxiously
pondering the situation thus created, a saintly old man was seen passing the
palace on his daily round of charity, followed by his servant carrying a load
of bread. It was deemed a happy omen, for the need of an understanding
with the Pope was urgent, and Zorzi il Santo would be an excellent mediator.
Being entreated, he accepted the charge and filled the ducal chair for ten
months, during which period he was able to obtain a relaxation of the
interdict, which was finally removed in the next reign.
Giorgio the Sainted left a troublesome legacy to his successor, Giov.
Soranza. Zara, aided by Hungarians and Croats, was again recalcitrant, and
only subdued after a heavy expenditure of men and treasure. But Soranza’s
sixteen and a half years of office coincided with a time of great prosperity,
and the strain was lightly borne. Venetian trade, aided by diplomacy and
enterprise, expanded eastward and westward. The arts of life were
developed. Refugees from Lucca founded a silk industry, which became a
source of great profit to the Republic. They were governed by their own
magistrates, the Provisores Sirici, who were located in the Corte della Seda,
near Marco Polo’s house. The city was further embellished, and Soranza
enjoyed the popularity that comes to a prince ruling in times of plenty. It was
in Soranza’s reign, August 1321, that Dante came to Venice, an ambassador
from Guido Novello da Polenta of Ravenna, to negotiate a peace with the
Signory, and returned to die a few days after his arrival at Ravenna of a fever
caught on the journey.

CHAPTER VIII

Conquests on the Mainland—Execution of Marin Faliero—The Fall of Genoa

“Ill-fated chief——
Him, only him, the shield of Jove defends
Whose means are fair and spotless as his ends.”
—Wordsworth.

WHEN the Great Council met in 1328 after Soranza’s death the oldest
member rose, and after uttering praises of the late Doge and lamentations at
his death, exhorted all around to be of good heart and to pray God for the
election of a wise prince to succeed him.
And never had Venice greater need of wisdom in her rulers. Unhappy
Italy, “reeling like a pilotless vessel in a mighty tempest,” had seen the last
vain attempt of the Emperors to execute their ideal function. The heroic
spirit of Henry VII. of Luxemburg, l’alto Arrigo, had already ascended to fill
the vacant and crowned seat, which Dante saw awaiting him among the
ranks of the blest, and the poet’s hopes of a Cæsar firmly seated in the saddle
and curbing the wanton and savage beast of faction in Italy were shattered.
From the political chaos three great families of despots emerged in North
Italy, the Scalas of Verona, the Viscontis of Milan and the Carraras of Padua.
In 1329 the lords of Verona under Mastino della Scala held sway over
Vicenza, Padua and Treviso. Mastino, an ambitious prince subtly urged by
the deposed Marsilio da Carrara of Padua, determined to tap the wealth of
the Republic by levying duties on all Venetian goods passing through his
territories. Venice retaliated by cutting off the supply of salt and a tariff war
began. But her food supplies soon ran short and an appeal to arms impended.
It was a tremendous choice, for the strength of Venice lay in her naval not in
her military power. To fight the mainland despots she must employ
mercenaries and successful condottieri were bad servants. Her infallible
bulwark, the sea, would be gone, squandered for a vulnerable land frontier
and the financial drain of a continental policy.
Such were the considerations that appealed to the mind of the new Doge
Francesco Dandolo. But the time was past when a Doge could dictate the
policy of the State. The oligarchy felt that inaction meant national suicide. If
the passage for Venetian goods to North and West Europe was blocked, it
meant ruin to her trade, and war was declared in the chivalrous fashion of
the times. To the confines of Padua, envoys were sent, who delivered a
protest and in token of defiance, three times cast a stone into the enemy’s
territory. A levy was made on all males between twenty and sixty years of
age, and alliances concluded with Florence, Milan, Ferrara and Mantua.
Mastino, alarmed at the coalition against him, sent Marsilio da Carrara to
treat with the Signory. Marsilio was invited, so the story runs, to sit next the
Doge at dinner. The “wily old fox” made a sign to the Doge that he wished
to speak with him. The Doge let fall his knife. Marsilio bent down to pick it
up and as the Doge bent down too, whispered: “What would you give him
who gave you Padua?” “We would make him lord of Padua,” replied the
Doge. Mastino was unable to stand before so powerful a combination. Padua
by collusion with the Carraras fell to Venetian arms, and Alberto della Scala,
Mastino’s brother, was led captive to Venice. Marsilio became once more
lord of Padua. The provinces of Treviso and Bassano were made over to
Venice, who thus came into possession of one of the passes into North
Europe and a rich corn land in North Italy. The glorious initiation of this new
and vaster policy was celebrated with great rejoicings at Venice and the
prophets of evil were silenced. The Republic proved herself a wise and
tolerant mistress. Her new subjects were ruled with a paternal regard for
their welfare, while local feeling and characteristics were tenderly treated.
The device which still remains from Venetian times over the town-hall of
Verona, Pro summa fide summus amor, admirably expresses the attitude of
the Republic towards her mainland provinces. The citizens of many a State
scourged with the scorpions of Italian despots hailed with delight the advent
of Venetian rule. Sanudo in his diary describes the entry into Faenza of the
Venetian governor on its occupation in 1495. The streets were decorated
with tapestry and cloth. For many days the painters had been doing nought
but paint San Marco on the doors of the houses. The whole city with great
demonstration of joy came forth to receive the new governor, shouting
“Marco! Marco!” A quarter of a century later the Venetian ambassador was
able to tell Cardinal Wolsey that although the Duke of Milan and the
Marquis of Mantua were candidates for the governorship of Verona, the
Venetian army as it defiled through the mountains was received by the
Veronese with joyful cries of “Marco! Marco!” and the Venetian
commissioner installed in the seat of government. St Mark’s lion on his
column standing in the market-place of an Italian city was the symbol of a
firm, just, enlightened rule.
While Bart. Gradenigo was Doge, Venice was visited by an awful
tempest. For three days the angry waves surged against the city. On the third
night, February 25, 1340, so runs the legend, as the storm increased in
violence a poor old fisherman was making fast his boat at the Molo when he
was accosted by a stranger who craved to be ferried over to the island of S.
Giorgio Maggiore. The fisherman, terrified by the awful storm, would not
venture, but being urged and reassured by the stranger, he loosed his boat
and they set forth and reached the island. Here a mysterious knight
embarked and commanded that they should be rowed to S. Nicolo on the
Lido, and on the boatman protesting the danger of going with one oar he was
again comforted and promised good pay. Arrived at S. Nicolo, a third
stranger, a venerable old man, joined them and demanded to be taken out to
sea. Scarcely had they reached the open Adriatic when they beheld a ship
filled with devils pressing swiftly forward to wreak destruction on Venice.
Soon as the three strangers sighted the vessel they made the sign of the cross
and ship and devils vanished. The sea grew calm: Venice was saved. The
three strangers were then rowed back whence they departed and revealed
themselves as St Mark, St George and St Nicholas. As St Mark stepped
ashore he was stayed by the boatman, who demurred to accepting the honour
of taking part in the miracle as adequate payment. “Thou art right,”
answered the saint, “go to the Doge and tell what thou hast seen and ask thy
reward. And say this has happened because of the master of a scuola at S.
Felice, who had sold his soul to the devil and at last hanged himself.” The
old man protested: “Even though I tell this, the Doge will not believe me.” St
Mark then drew a ring from his finger the worth of which was about five
ducats, gave it to the fisherman, saying, “Show this to the Doge and bid him
keep it in my sanctuary.” The saint’s bidding was done and the fisherman
well rewarded.
 In 1343 Andrea Dandolo was raised to the Dogeship. The first patrician
who had taken a doctor’s degree at Padua, he had already served on the Ten,
and filled the office of Procurator of St Mark, when at thirty-six years of age
he was elected Doge. Plague, earthquake and war scourged the Venetians
during the scholar Doge’s reign. Zara, tempted by the Hungarians, tried
another fall with her Venetian masters for independence. Marin Faliero, a
member of one of the most ancient and illustrious families of Venice, who
had served as podestà of Padua and of Treviso, was placed in command of a
land force: forty galleys set forth under Pietro da Canale to attack by sea.
Faliero met the Hungarians forty thousand strong about eight miles from
Zara, and won a decisive victory. Meanwhile Canale had forced the harbour,
and in a few months the tough old fortress again surrendered. The walls were
dismantled, and this time a strong Venetian garrison was left to overawe the
Zarantines.
For a century Genoese and Venetians had been rivals in the Crimea for
the control of the fur trade with Tartary. In 1346 Marin Faliero was sent to
Genoa to complain of certain piratical acts by the Genoese. The latter
retorted by accusing the Signory of bad faith in dealing with the Tartars. But
the ravages of the Black Death in 1348, when two-fifths of the people of
Venice are said to have perished, and fifty noble families to have become
extinct, absorbed for a time the attention of the Republic. In 1349 friction
with the Genoese led to a definite breach, and another act in the tragedy of
Christendom was begun. Two years of naval warfare passed with little
advantage to either power. In 1353, in alliance with Peter of Aragon and the
Emperor of Constantinople, the Venetians sailed up the Bosphorus and came
upon the enemy under Paganino Doria, off Pera. Two hours before sunset the
engagement began and a long and bloody struggle raged through the night,
illumined only by the lurid glare of burning ships. The Venetians were
defeated, and Nicolo Pisani, their commander, retired to Crete to recruit and
await reinforcements. In February 1353 he was able to join the Aragonese
fleet in Sardinian waters and attack the Genoese under Ant. Grimaldi off
Lojera. The enemy were outmanœuvred, their armament utterly destroyed,
and Grimaldi, with a few shattered galleys, reached Genoa crushed and
humiliated. Panic seized the city. In the despair engendered by defeat a
momentous step was taken. The cowed Genoese surrendered their
independence and implored the protection of Giov. Visconti, Lord Bishop of
Milan. Time for recuperation was, however, needed, and Visconti sought
peace for his new vassal by sending Petrarch to Venice as his envoy. The
poet had already, by an epistle to Dandolo, which reads like an echo of his
pathetic canzone All’Italia, besought the rivals to exchange the kiss of peace
and not persist in a war which must end in one of the eyes of Italy being
quenched, the other dimmed. As well appeal to two eagles fighting for their
quarry. After some weeks at Venice, honoured as a poet but unheeded as a
messenger of peace, he returned sorrowing to Milan. A year passed, and a
new fleet left Genoa under Doria, who cleverly slipped by Pisani, sailed up
the Adriatic, devastated Lesena, Curzola and Parenza, and anchored within
striking distance of the lagoons. It was now the turn of Venice to feel alarm.
The channels were fortified, a new fleet was equipped, a war tax levied. The
Doge, prostrated by the news, never recovered, and died of a broken heart on
September 7th, 1354. He lies in the Baptistery that he did so much to
beautify, and under his noble and simple monument there still remains the
Latin epitaph composed for him by Petrarch, his friend. Andrea Dandolo is
remembered as a Venetian historian and an accomplished legist, a lover of
the arts, and a great humanist.
On October the 4th the Bucintoro was sent to Chioggia to meet Marin
Faliero, recalled from the Roman legation to assume the ducal cap. The great
State barge set forth, but so dense a fog enveloped Venice that it was deemed
prudent to land in small boats. The gondola bearing Faliero failed to make
the usual stage by the Ponte della Paglia, and, sinister augury, landed the
Doge between the red columns of the Piazzetta.
Faliero, however, began his term of office happily by signing a truce with
the Genoese. The breathing time was used in concentrating the fleet, and
scarcely was the term ended when Nicolo Pisani, with his son Vettore, sailed
for the Ionian Isles in search of Doria. But the Genoese refused to be drawn,
and Pisani went into winter quarters at Portolungo in the Morea behind the
island of Sapienza. Doria saw an opportunity to trap his enemy. By a
brilliant manœuvre he got part of his fleet between the Venetians and the
shore, and attacking front and rear routed them with terrible loss, no less
than six thousand being made prisoners. Pisani and his lieutenant, Quirini,
escaped with the remnant of the armament to be impeached and degraded in
Venice. Under the shadow of this disaster Venice displayed her wonted
fortitude. With admirable self-control the Signory exhorted all their podestà,
consuls and agents to be of good courage, and called for men and ships in
defence of the fatherland. Eight thousand ducats were despatched to Genoa
to soften the lot of the captives. Before, however, the spring made further
operations possible, a truce was effected by the mediation of the Emperor;
and while negotiations were pending for a definite peace, Europe was
shocked by the news of the trial and execution of a Doge of Venice. The
chroniclers give very circumstantial details of the drama, but to Petrarch,
who was on terms of closest intimacy with the Doge, and to other
contemporaries the whole business was shrouded in mystery. No record of
the trial exists.
Faliero was of a proud and fiery temper. While podestà at Treviso he is
said to have boxed the bishop’s ears for having kept him waiting at a
religious procession. After the usual bull-fight and festivities in the Piazza on
Carnival Thursday the Doge gave a sumptuous banquet and dance in the
palace. Among the guests was a young noble, Michel Steno, who, heated
with wine, misconducted himself and was by the Doge’s orders expelled
from the hall. Steno, furious with rage, pencilled on the ducal chair, as he
passed through the Doge’s apartment, an insulting reflection on the Doge’s
honour: Marin Falier della bella mujer tu la mantien e altri la galde. Steno
was accused before the Quarantia and let off with a punishment which the
Doge regarded as derisory. On the morrow it befell that a choleric noble,
Marco Barbaro, went to demand certain things of the Arsenal authorities,
and being refused struck the Admiral Ghisello with his clenched fist. He was
wearing a ring and an ugly wound was left. The Admiral, with bloody face
went straightway to the Doge and prayed to be avenged. “What would you?”
said the Doge; “Look at the shameful words written of me and the way that
ribald Steno has been treated.” “Messer lo dose,” answered Ghisello, “if you
will make yourself lord of Venice and get all those cuckold nobles cut in
pieces I am prepared to help you.” The tempter spoke to willing ears. A vast
conspiracy was formed to make Faliero despot of Venice. Secret meetings
were held in the palace, and it was arranged that sixteen leaders should each
prepare an armed force of forty men, who, however, were not to be told for
what purpose they were wanted. On Wednesday, April 15th, a rumour was to
be spread that the Genoese were in the lagoons, and the alarm-bell of St
Mark’s to be rung. The leaders with their men were to converge on the
Piazza and cut down the nobles and influential citizens who would hasten to
answer the call. Faliero was then to be acclaimed lord of Venice. To incite
the people to hatred of the nobles, hired roysterers were sent by night about
the city doing violence to and insulting the women of the people, calling
each other by well-known names of nobles as they mockingly rioted through
the streets. “But the Lord God who hath ever watched over this most
glorious city,” says Sanudo, “inspired Beltrame Bergamasco, one of the
leaders, to go secretly to a beloved patron, Nicolo Lioni, and warn him as he
valued his life not to leave his house on the 15th.” Lioni’s suspicions were
aroused; he questioned Beltrame and finally locked him in a room while he
went to seek Giov. Gradenigo, dubbed Nasone (big nose) and Marco
Cornaro in order to take counsel together. They returned to Lioni’s house
and after again questioning Beltrame decided to send messengers with an
urgent summons to the Ten and the chief Councillors and officers of the
State to a meeting in the sacristy of S. Salvatore. Blanched with terror they
hurried together. Up to that time the complicity of the Doge was not known,
but the immediate arrest of two other conspirators discovered the full danger
of the situation. The Ten were equal to the emergency. The chief police
officials and heads of the wards were sent for and asked to take good men
and true, and arrest the leading conspirators. Armed forces were summoned
from Chioggia. All means of egress from the palace were guarded; the bell-
ringer forbidden to stir. On the dawn of the 16th the Ten summoned a Zonta
of twenty of the wisest, best and most ancient of the nobles of Venice to join
them in trying the Doge. He made no attempt to defend himself and was
sentenced to death. On the 17th, the biretta removed, stripped of his ducal
robes and clothed in a black gown, he was led to the spot on the landing of
the staircase[39] in the courtyard where seven months before he had sworn to
defend the Constitution. At one stroke his head was severed from his body. A
capo of the Ten then went to the arcade of the palace over the Piazza and
showed the executioner’s bloody sword to the assembled people, crying:
“The great judgment has been done on the traitor.” The gates were opened
and the people rushed in to see the body. At night all that remained of Marin
Faliero was placed in a boat with eight lighted torches and carried to burial
at S. Zanipolo. No record of the sentence is found in the acts of the Ten. On
the unfilled space where the minutes should be entered are the words Ñ.
Sc̄ batur, “Let it not be written.” A year later his portrait was blotted out and
the place covered with a black veil with the inscription: Hic est locus Marini
Faletro decapitati pro criminibus. Meanwhile the leaders of the plot as they
were captured were hanged in a row, with iron gags in their mouths, from the
columns of the palace. St Isidore’s Day, April 16th, was made a public
festival and as late as 1520 Sanudo saw carried in procession the white
damask cloth bespattered with blood which was used at the execution.

THE PALAZZI GIUSTINIANI AND FOSCARI

On the 21st the vacant chair was filled by the election of Gradenigo il
nasone, who had been largely instrumental in frustrating the plot. The
Genoese negotiations dragged on, and not until the June following was the
treaty of peace signed at Milan. Venetian statesmen had short rest from
foreign complications. The defeat of Sapienza and the supposed laming of
the executive by the miserable end of Doge Faliero, tempted the Hungarians,
who were rapidly increasing in population and civilisation, to achieve their
purpose of acquiring possession of a sea-board. War was declared and
Francesco Carrara, who held Padua under Venetian tutelage, was called to
aid his suzerain. But the Carraras aimed at founding a dynasty and gave
secret assistance to King Louis of Hungary, who was besieging Treviso. The
defence of Treviso and of Dalmatia was beyond the power of the Republic.
After a two years’ struggle and bitter hours of humiliation, Dalmatia, bought
with so much blood of Doges, patricians and people, was surrendered by the
peace of Zara, in February 1358, to the King of Hungary in exchange for the
retention of Treviso. Meantime Gradenigo had died (in 1356) and left to
Giov. Delfino the signing of the great renunciation. Delfino was stoutly
defending Treviso when elected, and being refused a permit to pass the
Hungarian lines broke through by night and met the Senate at Mestre.
Broken-hearted, his sight failing, he died of the plague in 1361. Under
Delfino’s successor Lorenzo Celsi, Crete, never wholly subjugated, burst
into revolt and terror reigned in the island. At this time Petrarch, seeking
peace and security, had settled in Venice, the “only nest of liberty and sole
refuge of the good.” He lived simply, at the sign of the Two Towers on the
Riva degli Schiavoni, where it was his delight to stand at his window
watching the huge galleys, big as the house he lived in, and with masts taller
than towers, passing and repassing. The poet, honoured and cherished, was
on intimate terms with the Signory and advised the employment of the
famous Veronese condottiero Lucchino del Verme to subdue the island. On
the 4th of June about the sixth hour the poet was at his window chatting with
the Archbishop of Patras when the friends saw a galley gliding swiftly up the
lagoons, her masts wreathed with flowers; her deck crowded with men
waving flags; hostile banners trailed in the waters behind her. She brought
the news that Lucchino, falling upon the insurgents, weakened by divided
counsels, had defeated them and punished the leaders. A thanksgiving
festival was ordered which lasted three days. “The Doge,” writes Petrarch,
“with a numerous train took his place to watch the sports over the vestibule
of St Mark’s, where stand the four bronze horses (a work of ancient and
excellent art) that seem to challenge comparison with the living and raise
their hoofs to tread the ground. An awning of tapestry in many colours kept
off the heat of the sun and I myself was invited to sit by the right hand of the
Doge.” The great Piazza, the church, the Campanile, the roofs, the arcades,
the windows, were a mass of people. A magnificent pavilion next the church
was filled with more than four hundred gaily dressed ladies.
English knights took part in the jousts. The poet, bored by the length of
the festivities, pleaded pressure of business on the second day and came no
more. But the rejoicings were premature, the revolt soon flamed forth again,
and a costly campaign of twelve months was needed to quench it. Petrarch,
to show his gratitude to the Venetian State, offered to make over his great
collection of books to found a public library if the Republic would house
them. The procurators of St Mark accepted the charge, but the ultimate fate
of this priceless gift is unknown. During the short and peaceful reign of
Marco Corner, Guariento of Verona was employed to paint on the walls of
the Hall of the Great Council the story of Alexander III. and Barbarossa, for
which Petrarch composed inscriptions. In 1368 a deputation from the Great
Council headed by Vettor Pisani went in search of the procurator, Andrea
Contarini, who had been chosen Doge, and found him grafting fruit trees on
his farm by the Brenta. Contarini had been warned against accepting the
Dogeship by a Syrian sooth-sayer, and threats of confiscation were necessary
to force him from his retreat.
At the peace of Zara, Louis of Hungary had shielded Carrara from
Venetian vengeance, but the erection of two forts on the Brenta now gave the
Venetians a pretext for paying off old scores. Louis again stood by his ally
until the lucky capture of the King’s nephew by the Venetians enabled them
to exact the abandonment of the war as the price of his ransom, and Carrara
was made to pay heavily for disloyalty. Petrarch accompanied the Paduan
peace envoys—a grateful task to the poet of peace and concord.[40] It was
his last mission to Venice.
The final act in the struggle with Genoa now draws nigh. A quarrel for
precedence took place at the Coronation of the King of Cyprus between the
Venetian and Genoese envoys, Malipiero and Paganino Doria, in which the
latter, “being full of anger and venom,” made use of coarse and unseemly
words towards the Venetians. The quarrel was renewed at the banquet that
followed, and loaves of bread and other meats were used as missiles. The
Cypriotes sided with the Venetians, and many of the Genoese present were
flung out of the window and dashed to pieces. But it was a dispute for the
possession of the classic Tenedos notissima famæ insula which made war
inevitable. The Signory offered the Greek Emperor for the cession of the
island (which by its position south of the Dardanelles was of great strategical
importance) the sum of three thousand ducats and the return of the Imperial
jewels which were held in pledge at Venice. The offer was made a demand
by a threat to treat with the Turks if the Emperor refused the bargain.
Paleologus had emptied his treasury to meet the cost of the Ottoman wars,
and accepted the offer. The Genoese retaliated by taking in hand Andronicus,
the Emperor’s rebellious and renegade son, who in return for their support
promised to make them masters of the island. The islanders confronted by
the rival claims, came forth bearing crosses in their hands to welcome the
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!

textbookfull.com