Data-Driven Science and Engineering Steven L.

Brunton pdf download

https://ebookmeta.com/product/data-driven-science-and-
engineering-steven-l-brunton/

Download more ebook from https://ebookmeta.com

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

Data driven science and engineering 2nd Edition Steven

Brunton

https://ebookmeta.com/product/data-driven-science-and-
engineering-2nd-edition-steven-brunton/

Data-Driven Engineering Design 2022nd Edition Liu

https://ebookmeta.com/product/data-driven-engineering-
design-2022nd-edition-liu/

Data-Driven Evolutionary Optimization: Integrating

Evolutionary Computation, Machine Learning and Data
Science (Studies in Computational Intelligence, 975)
Yaochu Jin
https://ebookmeta.com/product/data-driven-evolutionary-
optimization-integrating-evolutionary-computation-machine-
learning-and-data-science-studies-in-computational-
intelligence-975-yaochu-jin/

The Waistcoat Workbook: Historical, Modern, and Genre

Drafting of Waistcoats for Men and Women 1837–Present
Day 1st Edition J. François-Campbell

https://ebookmeta.com/product/the-waistcoat-workbook-historical-
modern-and-genre-drafting-of-waistcoats-for-men-and-
women-1837-present-day-1st-edition-j-francois-campbell/
Geology of the World’s Major Gold Deposits and
Provinces R.H. Sillitoe

https://ebookmeta.com/product/geology-of-the-worlds-major-gold-
deposits-and-provinces-r-h-sillitoe/

Art History For Dummies Jesse Bryant Wilder

https://ebookmeta.com/product/art-history-for-dummies-jesse-
bryant-wilder/

Leading through Disruption: A Changemaker's Guide to

Twenty-First Century Leadership 1st Edition Andrew
Liveris

https://ebookmeta.com/product/leading-through-disruption-a-
changemakers-guide-to-twenty-first-century-leadership-1st-
edition-andrew-liveris/

BDD in Action, 2nd Edition (Final Release) John

Ferguson Smart

https://ebookmeta.com/product/bdd-in-action-2nd-edition-final-
release-john-ferguson-smart/

Sex Work and Hate Crime Rosie Campbell

https://ebookmeta.com/product/sex-work-and-hate-crime-rosie-
campbell/
The Fairness Doctrine and the Media Steven J. Ssmmons

https://ebookmeta.com/product/the-fairness-doctrine-and-the-
media-steven-j-ssmmons/
Data-Driven Science and Engineering

Data-driven discovery is revolutionizing how we model, predict, and control complex

systems. Now with Python and MATLAB , this textbook trains mathematical scientists
and engineers for the next generation of scientific discovery by offering a broad overview of
the growing intersection of data-driven methods, machine learning, applied optimization,
and the classic fields of engineering mathematics and mathematical physics. With a
focus on integrating dynamical systems modeling and control with modern methods in
applied machine learning, this text includes methods that were chosen for their relevance,
simplicity, and generality. Topics range from introductory to research-level material,
making it accessible to advanced undergraduate and beginning graduate students from
the engineering and physical sciences. This second edition features new chapters on
reinforcement learning and physics-informed machine learning, significant new sections
throughout, and chapter exercises. Online supplementary material – including lecture
videos per section, homework, data, and codes in MATLAB , Python, and R – is available
on http://databookuw.com.

Steven L. Brunton is the James B. Morrison Professor of Mechanical Engineering at the

University of Washington (UW) and Associate Director of the NSF AI Institute in Dynamic
Systems. He is also Adjunct Professor of Applied Mathematics and Computer Science,
and a Data-Science Fellow at the eScience Institute. His research merges machine learning
with dynamical systems and control, with applications in fluid dynamics, bio-locomotion,
optics, energy systems, and manufacturing. He is an author of three textbooks, and received
the UW College of Engineering Teaching Award, the Army and Air Force Young Investi-
gator Program (YIP) Awards, and the Presidential Early Career Award for Scientists and
Engineers (PECASE).

J. Nathan Kutz is the Robert Bolles and Yasuko Endo Professor of Applied Mathematics
at the University of Washington (UW) and Director of the NSF AI Institute in Dynamic
Systems. He is also Adjunct Professor of Electrical and Computer Engineering, Mechanical
Engineering, and Physics, and Senior Data-Science Fellow at the eScience Institute. His
research interests lie at the intersection of dynamical systems and machine learning. He is
an author of three textbooks and has received the Applied Mathematics Boeing Award of
Excellence in Teaching and an NSF CAREER Award.
“Finally, a book that introduces data science in a context that will make any mechanical
engineer feel comfortable. Data science is the new calculus, and no engineer should grad-
uate without a thorough understanding of the topic”
Hod Lipson, Columbia University

“This book is a must-have for anyone interested in data-driven modeling and simulations.
Readers as diverse as undergraduate STEM students and seasoned researchers would find
it useful as a guide to this rapidly evolving field. Topics covered by the monograph include
dimension reduction, machine learning, and robust control of dynamical systems with
uncertain/random inputs. Every chapter contains codes and homework problems, which
make this treatise ideal for the classroom setting. The book is supplemented with online
lectures, which are not only educational but also entertaining to watch.”
Daniel M. Tartakovsky, Stanford University

“Engineering principles will always be based on physics, and the models that underpin
engineering will be derived from these physical laws. But, in the future, models based on
relationships in large datasets will be as important and, when used alongside physics-based
models, will lead to new insights and designs. Brunton and Kutz will equip students and
practitioners with the tools they will need for this exciting future.”
Greg Hyslop, Boeing

“Brunton and Kutz’s book is fast becoming an indispensable resource for machine learn-
ing and data-driven learning in science and engineering. The second edition adds several
timely topics in this lively field, including reinforcement learning and physics-informed
machine learning. The text balances theoretical foundations and concrete examples with
code, making it accessible and practical for students and practitioners alike.”
Tim Colonius, California Institute of Technology

“This is a must-read for those who are interested in understanding what machine learning
can do for dynamical systems! Steve and Nathan have done an excellent job in bringing
everyone up to speed with the modern application of machine learning to these complex
dynamical systems.”
Shirley Ho, Flatiron Institute and New York University
Data-Driven Science and
Engineering
Machine Learning, Dynamical Systems,
and Control
Second Edition

STEVEN L. BRUNTON
University of Washington

J . N AT H A N K U T Z
University of Washington
University Printing House, Cambridge CB2 8BS, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,
New Delhi – 110025, India
103 Penang Road, #05–06/07, Visioncrest Commercial, Singapore 238467

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.

www.cambridge.org
Information on this title: www.cambridge.org/highereducation/isbn/9781009098489
DOI: 10.1017/9781009089517
© Steven L. Brunton and J. Nathan Kutz 2022
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2019
Second edition 2022
Printed in the United Kingdom by TJ Books Limited, Padstow, Cornwall
A catalogue record for this publication is available from the British Library.
ISBN 978-1-009-09848-9 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy
of URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
 Contents

Preface page ix
Acknowledgments xiv
Common Optimization Techniques, Equations, Symbols, and Acronyms xv

Part I Dimensionality Reduction and Transforms 1

1 Singular Value Decomposition (SVD) 3

1.1 Overview 3
1.2 Matrix Approximation 7
1.3 Mathematical Properties and Manipulations 12
1.4 Pseudo-Inverse, Least-Squares, and Regression 16
1.5 Principal Component Analysis (PCA) 23
1.6 Eigenfaces Example 28
1.7 Truncation and Alignment 35
1.8 Randomized Singular Value Decomposition 40
1.9 Tensor Decompositions and N-Way Data Arrays 46

2 Fourier and Wavelet Transforms 53

2.1 Fourier Series and Fourier Transforms 53
2.2 Discrete Fourier Transform (DFT) and Fast Fourier Transform (FFT) 63
2.3 Transforming Partial Differential Equations 70
2.4 Gabor Transform and the Spectrogram 76
2.5 Laplace Transform 81
2.6 Wavelets and Multi-Resolution Analysis 85
2.7 Two-Dimensional Transforms and Image Processing 87

3 Sparsity and Compressed Sensing 97

3.1 Sparsity and Compression 97
3.2 Compressed Sensing 101
3.3 Compressed Sensing Examples 105
3.4 The Geometry of Compression 109
3.5 Sparse Regression 113
3.6 Sparse Representation 117
3.7 Robust Principal Component Analysis (RPCA) 120
3.8 Sparse Sensor Placement 123

v
 vi Contents

Part II Machine Learning and Data Analysis 131

4 Regression and Model Selection 133

4.1 Classic Curve Fitting 134
4.2 Nonlinear Regression and Gradient Descent 140
4.3 Regression and Ax = b: Over- and Under-Determined Systems 145
4.4 Optimization as the Cornerstone of Regression 151
4.5 The Pareto Front and Lex Parsimoniae 155
4.6 Model Selection: Cross-Validation 158
4.7 Model Selection: Information Criteria 162

5 Clustering and Classification 168

5.1 Feature Selection and Data Mining 169
5.2 Supervised versus Unsupervised Learning 174
5.3 Unsupervised Learning: k-Means Clustering 178
5.4 Unsupervised Hierarchical Clustering: Dendrogram 182
5.5 Mixture Models and the Expectation-Maximization Algorithm 186
5.6 Supervised Learning and Linear Discriminants 189
5.7 Support Vector Machines (SVM) 193
5.8 Classification Trees and Random Forest 198
5.9 Top 10 Algorithms of Data Mining circa 2008 (Before the Deep
Learning Revolution) 203

6 Neural Networks and Deep Learning 208

6.1 Neural Networks: Single-Layer Networks 209
6.2 Multi-Layer Networks and Activation Functions 214
6.3 The Backpropagation Algorithm 219
6.4 The Stochastic Gradient Descent Algorithm 222
6.5 Deep Convolutional Neural Networks 224
6.6 Neural Networks for Dynamical Systems 228
6.7 Recurrent Neural Networks 233
6.8 Autoencoders 236
6.9 Generative Adversarial Networks (GANs) 240
6.10 The Diversity of Neural Networks 242

Part III Dynamics and Control 251

7 Data-Driven Dynamical Systems 253

7.1 Overview, Motivations, and Challenges 254
7.2 Dynamic Mode Decomposition (DMD) 260
7.3 Sparse Identification of Nonlinear Dynamics (SINDy) 275
7.4 Koopman Operator Theory 286
7.5 Data-Driven Koopman Analysis 296
 Contents vii

8 Linear Control Theory 311

8.1 Closed-Loop Feedback Control 312
8.2 Linear Time-Invariant Systems 317
8.3 Controllability and Observability 322
8.4 Optimal Full-State Control: Linear–Quadratic Regulator (LQR) 328
8.5 Optimal Full-State Estimation: the Kalman Filter 332
8.6 Optimal Sensor-Based Control: Linear–Quadratic Gaussian (LQG) 335
8.7 Case Study: Inverted Pendulum on a Cart 336
8.8 Robust Control and Frequency-Domain Techniques 346

9 Balanced Models for Control 360

9.1 Model Reduction and System Identification 360
9.2 Balanced Model Reduction 361
9.3 System Identification 375

Part IV Advanced Data-Driven Modeling and Control 387

10 Data-Driven Control 389

10.1 Model Predictive Control (MPC) 390
10.2 Nonlinear System Identification for Control 392
10.3 Machine Learning Control 398
10.4 Adaptive Extremum-Seeking Control 408

11 Reinforcement Learning 419

11.1 Overview and Mathematical Formulation 419
11.2 Model-Based Optimization and Control 426
11.3 Model-Free Reinforcement Learning and Q-Learning 429
11.4 Deep Reinforcement Learning 436
11.5 Applications and Environments 440
11.6 Optimal Nonlinear Control 444

12 Reduced-Order Models (ROMs) 449

12.1 Proper Orthogonal Decomposition (POD) for Partial Differential Equations 449
12.2 Optimal Basis Elements: the POD Expansion 455
12.3 POD and Soliton Dynamics 461
12.4 Continuous Formulation of POD 465
12.5 POD with Symmetries: Rotations and Translations 470
12.6 Neural Networks for Time-Stepping with POD 475
12.7 Leveraging DMD and SINDy for Galerkin–POD 479

13 Interpolation for Parametric Reduced-Order Models 485

13.1 Gappy POD 485
13.2 Error and Convergence of Gappy POD 490
13.3 Gappy Measurements: Minimize Condition Number 493
13.4 Gappy Measurements: Maximal Variance 497
 viii Contents

13.5 POD and the Discrete Empirical Interpolation Method (DEIM) 500
13.6 DEIM Algorithm Implementation 504
13.7 Decoder Networks for Interpolation 508
13.8 Randomization and Compression for ROMs 512
13.9 Machine Learning ROMs 513

14 Physics-Informed Machine Learning 520

14.1 Mathematical Foundations 520
14.2 SINDy Autoencoder: Coordinates and Dynamics 523
14.3 Koopman Forecasting 526
14.4 Learning Nonlinear Operators 529
14.5 Physics-Informed Neural Networks (PINNs) 533
14.6 Learning Coarse-Graining for PDEs 535
14.7 Deep Learning and Boundary Value Problems 539

Glossary 542
References 552
Index 588
Preface

This book is about the growing intersection of data-driven methods, machine learning,
applied optimization, and the classical fields of engineering mathematics and mathematical
physics. We developed this material over a number of years, primarily to educate our
advanced undergraduate and beginning graduate students from engineering and physical
science departments. Typically, such students have backgrounds in linear algebra, differ-
ential equations, and scientific computing, with engineers often having some exposure to
control theory and/or partial differential equations. However, most undergraduate curricula
in engineering and science fields have little or no exposure to data methods and/or opti-
mization. Likewise, computer scientists and statisticians have little exposure to dynamical
systems and control. Our goal is to provide a broad entry point to applied machine learning
for both of these groups of students. We have chosen the methods discussed in this book
for their (1) relevance, (2) simplicity, and (3) generality, and we have attempted to present
a range of topics, from basic introductory material up to research-level techniques.
Data-driven discovery is currently revolutionizing how we model, predict, and control
complex systems. The most pressing scientific and engineering problems of the modern era
are not amenable to empirical models or derivations based on first principles. Increasingly,
researchers are turning to data-driven approaches for a diverse range of complex systems,
such as turbulence, the brain, climate, epidemiology, finance, robotics, and autonomy.
These systems are typically nonlinear, dynamic, multi-scale in space and time, and high-
dimensional, with dominant underlying patterns that should be characterized and modeled
for the eventual goal of sensing, prediction, estimation, and control. With modern math-
ematical methods, enabled by the unprecedented availability of data and computational
resources, we are now able to tackle previously unattainable problems. A small handful of
these new techniques include robust image reconstruction from sparse and noisy random
pixel measurements, turbulence control with machine learning, optimal sensor and actuator
placement, discovering interpretable nonlinear dynamical systems purely from data, and
reduced-order models to accelerate the optimization and control of systems with complex
multi-scale physics.
Driving modern data science is the availability of vast and increasing quantities of data,
enabled by remarkable innovations in low-cost sensors, orders-of-magnitude increases in
computational power, and virtually unlimited data storage and transfer capabilities. Such
vast quantities of data are affording engineers and scientists across all disciplines new
opportunities for data-driven discovery, which has been referred to as the fourth para-
digm of scientific discovery [325]. This fourth paradigm is the natural culmination of the

ix
x Preface

first three paradigms: empirical experimentation, analytical derivation, and computational

investigation. The integration of these techniques provides a transformative framework for
data-driven discovery efforts. This process of scientific discovery is not new, and indeed
mimics the efforts of leading figures of the scientific revolution: Johannes Kepler (1571–
1630) and Sir Isaac Newton (1642–1727). Each played a critical role in developing the
theoretical underpinnings of celestial mechanics, based on a combination of empirical data-
driven and analytical approaches. Data science is not replacing mathematical physics and
engineering, but is instead augmenting it for the twenty-first century, resulting in more of
a renaissance than a revolution.
Data science itself is not new, having been proposed more than 50 years ago by John
Tukey, who envisioned the existence of a scientific effort focused on learning from data,
or data analysis [204]. Since that time, data science has been largely dominated by two
distinct cultural outlooks on data [109]. The machine learning community, which pre-
dominantly comprises computer scientists, is typically centered on prediction quality and
scalable, fast algorithms. Although not necessarily in contrast, the statistical learning com-
munity, often centered in statistics departments, focuses on the inference of interpretable
models. Both methodologies have achieved significant success and have provided the math-
ematical and computational foundations for data science methods. For engineers and scien-
tists, the goal is to leverage these broad techniques to infer and compute models (typically
nonlinear) from observations that correctly identify the underlying dynamics and general-
ize qualitatively and quantitatively to unmeasured parts of phase, parameter, or application
space. Our goal in this book is to leverage the power of both statistical and machine learning
to solve engineering problems.

Themes of This Book

There are a number of key themes that have emerged throughout this book. First, many
complex systems exhibit dominant low-dimensional patterns in the data, despite the rapidly
increasing resolution of measurements and computations. This underlying structure enables
efficient sensing, and compact representations for modeling and control. Pattern extraction
is related to the second theme of finding coordinate transforms that simplify the system.
Indeed, the rich history of mathematical physics is centered around coordinate transfor-
mations (e.g., spectral decompositions, the Fourier transform, generalized functions, etc.),
although these techniques have largely been limited to simple idealized geometries and
linear dynamics. The ability to derive data-driven transformations opens up opportunities
to generalize these techniques to new research problems with more complex geometries
and boundary conditions. We also take the perspective of dynamical systems and control
throughout the book, applying data-driven techniques to model and control systems that
evolve in time. Perhaps the most pervasive theme is that of data-driven applied optimiza-
tion, as nearly every topic discussed is related to optimization (e.g., finding optimal low-
dimensional patterns, optimal sensor placement, machine learning optimization, optimal
control, etc.). Even more fundamentally, most data is organized into arrays for analysis,
where the extensive development of numerical linear algebra tools from the early 1960s
onward provides many of the foundational mathematical underpinnings for matrix decom-
positions and solution strategies used throughout this text.
 Preface xi

Overview of Second Edition

The integration of machine learning methods in science and engineering has advanced
significantly in the two years since publication of the first edition. The field is fast-moving,
with innovations coming in a diversity of application areas that use creative mathematical
architectures for advancing the state of the art in data-driven modeling and control. This
second edition is aimed at capturing some of the more salient and successful advancements
in the field. It helps bring the reader to a modern understanding of what is possible using
machine learning in science and engineering. As with the first edition, extensive online
supplementary material can be found at the book’s website:
http://databookuw.com
Major changes in the second edition include the following.

• Homework: Extensive homework has been added to every chapter, with additional
homework and projects on the book’s website. Homework ranges in difficulty from
introductory demonstrations and concept-building to advanced problems that repro-
duce modern research papers and may be the basis of course projects.
• Code: Python code has been added throughout, in parallel to existing MATLAB
code, and both sets of codes have been streamlined considerably. All extended codes
are available in MATLAB and Python on the book’s website and GitHub pages.
– Python Code:
https://github.com/dynamicslab/databook_python
– MATLAB Code:
https://github.com/dynamicslab/databook_matlab
Wherever possible, a minimal representation of code has been presented in the text to
improve readability. These code blocks are equivalently expressed in MATLAB and
Python. In more advanced examples, it is often advantageous to use either MATLAB
or Python, but not both. In such cases, this has been indicated and only a single code
block is demonstrated. The full code is available at the above GitHub sites as well
as on the book’s website. In addition, extensive codes are available in R online. We
encourage the reader to read the book and follow along with code to help improve
the learning process and experience.
• New chapters: Two new chapters have been added on “Reinforcement Learning”
and “Physics-Informed Machine Learning,” which are two of the most exciting and
rapidly growing fields of research in machine learning, modeling, and control.
– Reinforcement Learning: Reinforcement learning is a third major branch of
machine learning that is concerned with how to learn control laws and policies
to interact with a complex environment. This is a critical area of research,
situated at the growing intersection of control theory and machine learning.
– Physics-Informed Machine Learning: The integration of physics concepts,
constraints, and symmetries is providing exceptional opportunities for training
machine learning algorithms that are encoded with knowledge of physics. This
chapter features a number of recent innovations aimed at understanding how
this can be done in principle and in practice.
xii Preface

• New sections: We have added and improved material throughout, including the
following.
– Chapter 1: new sections discussing condition number, connections to the
eigendecomposition, and error bounds for SVD (singular value decomposi-
tion) based approximations.
– Chapter 2: new section on the Laplace transform.
– Chapter 6: new sections devoted to autoencoders, recurrent neural networks,
and generative adversarial networks.
– Chapter 7: addition of recent innovations to DMD (dynamic mode decom-
position), Koopman theory, and SINDy (sparse identification of nonlinear
dynamics).
– Chapter 10: new section on model predictive control.
– Chapter 12 (previously Chapter 11): new sections on using neural networks
for time-stepping in reduced-order models, as well as non-intrusive methods
such as DMD.
– Chapter 13 (previously Chapter 12): new sections on decoder networks for
interpolation in model reduction as well as randomized linear algebra methods
for scalable reduced-order models.

• Videos: An extensive collection of video lectures are available on YouTube, covering

nearly every topic from each section of the book. Videos may be found on our
YouTube channels.
– www.youtube.com/c/eigensteve
– www.youtube.com/c/NathanKutzAMATH
– www.youtube.com/c/PhysicsInformedMachineLearning

• Typos: We have corrected typos and mistakes throughout the second edition.

Online Material
We have designed this book to make extensive use of online supplementary material,
including codes, data, videos, homework, and suggested course syllabi. All of this material
can be found at the book’s website: http://databookuw.com.
In addition to course resources, all of the code and data used in the book are available
on the book’s GitHub: https://github.com/dynamicslab/. The codes online are more exten-
sive than those presented in the book, including code used to generate publication-quality
figures. In addition to the Python and MATLAB used throughout the text, online code is
also available in R. Data visualization was ranked as the top-used data science method
in the Kaggle 2017 The State of Data Science and Machine Learning study, and so we
highly encourage readers to download the online codes and make full use of these plotting
commands.
We have also recorded and posted video lectures on YouTube for every section in this
book, available at www.youtube.com/c/eigensteve and www.youtube.com/c/NathanKutz
AMATH. We include supplementary videos for students to fill in gaps in their background
on scientific computing and foundational applied mathematics. We have designed this text
to be both a reference as well as the material for several courses at various levels of student
 Preface xiii

preparation. Most chapters are also modular, and may be converted into stand-alone boot
camps, containing roughly 10 hours of materials each.

How to Use This Book

Our intended audience includes beginning graduate students, or advanced undergraduates,
in engineering and science. As such, the machine learning methods are introduced at a
beginning level, whereas we assume students know how to model physical systems with
differential equations and simulate them with solvers such as ode45. The diversity of topics
covered thus range from introductory to state-of-the-art research methods. Our aim is
to provide an integrated viewpoint and mathematical toolset for solving engineering and
science problems. Alternatively, the book can also be useful for computer science and
statistics students, who often have limited knowledge of dynamical systems and control.
Various courses can be designed from this material, and several example syllabi may be
found on the book’s website – this includes homework, data sets, and code.
First and foremost, we want this book to be fun, inspiring, eye-opening, and empowering
for young scientists and engineers. We have attempted to make everything as simple as
possible, while still providing the depth and breadth required to be useful in research. Many
of the chapter topics in this text could be entire books in their own right, and many of them
are. However, we also wanted to be as comprehensive as may be reasonably expected for
a field that is so big and moving so fast. We hope that you enjoy this book, master these
methods, and change the world with applied data science!
Another Random Scribd Document
with Unrelated Content
 CHAPTER XVII.
MR. PACKARD OF COLORADO.

In a short time the Bentons were settled in a neat flat located near
Washington Square. They purchased additional furniture on the
installment plan, and were able to offer Rupert a home more
desirable than the room he had occupied. The new prosperity was
reflected in the faces of the now happy wife and mother.
"It is you who have brought this happy change in our circumstances,
Rupert," she said. "I tremble to think what would have been our
condition but for you."
"In return you give me a pleasant home," said Rupert.
At the hotel things went on pleasantly. Rupert's services were
appreciated, and this was pleasant, though his salary had not been
increased.
Clarence Clayton never entered the hotel now. Rupert wondered
what had become of him. But one Thursday afternoon—his
afternoon off—he strayed down to the Battery.
Seated on one of the benches, looking out towards Governor's
Island, Rupert's attention was drawn after a while to two men who
occupied a neighboring bench. One of those he recognized at once
as Clayton. The other he also recognized as a guest at the Somerset
Hotel, a new arrival. He was a man of middle age who had the
appearance of a Westerner. Rupert now remembered that he had
entered himself on the hotel register as from Colorado.
"I wonder what mischief Clayton is up to now?" Rupert asked
himself.
The benches were so near that he was able to hear the conversation
between the two men. Clayton had a showy gold watch in his hand
which he was endeavoring to sell to his new acquaintance.
"The fact is, my friend," Rupert heard him say, "I am awfully hard
up. I need money badly, and that is why I offer you such a bargain.
This watch is nearly new and cost me one hundred and fifty dollars
in cold cash. I offer it to you for fifty."
"How did you get so hard up?" asked the stranger.
"I took a flyer in Wall Street. I have a friend who is a broker, and he
gave me a pointer. I don't blame him, for he believed it, and
invested himself. However, things didn't turn out as we expected,
and I was cleaned out."
"How about him?"
"He lost a good deal more than I did, but he could stand it and I
couldn't."
The Western man took the watch in his hand.
"It seems a good watch," he said. "I suppose it is solid gold?"
"Undoubtedly."
"I don't know much about watches myself, though I come from a
mineral producing State. We have plenty of miners there, but I am a
cattleman."
"Indeed! Is that a paying business?"
"Well, I've made a little money at it," said the other in a complacent
tone.
"I am looking for a paying business myself."
The stranger laughed.
"You are a city man," he said. "You wouldn't do for the West. You
wouldn't make much of a cowboy."
"I don't suppose I should."
"You couldn't wear patent-leather shoes in Colorado."
"Then I'll give it up if you say so. To tell the truth, I am better fixed
than you would suppose. I have an income of a thousand dollars a
year, paid me quarterly by the trustees of my late uncle's estate, but
the next payment won't come due for a month. I must tide over till
then. That is why I offer you this watch for fifty dollars."
"I shouldn't think you would like to make such a sacrifice."
"Oh, well, I need the money. Besides, what is my loss is your gain."
"You seem to take matters philosophically."
"That's my way. Seriously, though, it will be a great favor to me if
you take the watch. Fifty dollars isn't much, but with economy it will
carry me through till my next payment."
"Well, if you put it on that ground, I don't know but I will oblige
you."
The Colorado man took from his pocket a large wallet, evidently
stuffed with bills, and was about to consummate the bargain when
Rupert rose from his seat hastily. He felt that it was about time for
him to take a hand in the transaction.
"Mr. Packard," he said, "you'll excuse my interfering, but I advise you
not to buy that watch."
Clarence Clayton looked up quickly. He recognized Rupert only too
well, and would liked to have pitched him into the bay. What was to
be done? He determined to brazen it out.
"Young man," he said sharply, "you'd better mind your own
business."
"How do you know my name?" asked the man from Colorado, not
recognizing Rupert.
"I am one of the bell-boys at the Somerset Hotel, where you are
boarding."
"Why do you give this warning? Can you judge of the value of the
watch?"
"No, sir; but I know this man."
"That is false," asserted Clayton; "I never saw you before to my
knowledge."
"I don't know what to think," said the cattleman, looking puzzled.
"You say you know this man?"
"Yes. He came near cheating one of our guests not long since by
offering to give his son a place in an office in Wall Street for a
hundred and fifty dollars."
"The boy lies," exclaimed Clayton. "I have a good mind to give you
in charge, you young rascal."
"You are quite welcome to do it," said Rupert, coolly.
"I hope my word is as good as this boy's," continued Clayton.
"Don't take either, Mr. Packard. I am no judge of watches. Suppose
you go to a jeweler's and ask him the value of it. If it is worth even
a hundred dollars, you can venture to give this man what he asks,
that is, supposing he has come by it honestly."
"That is a sensible proposal. I accept it."
"But I don't!" said Clayton. "I feel that I have been insulted, and I
decline to sell the watch. As for you, you young rascal, I shall
remember your interference with me in my business."
He rose and went off with his head very high in the air.
"Sit down and tell me all about this fellow," said the cattleman. "I
suspect you have saved me from being imposed upon."
Rupert told the story, and the stranger thanked him heartily.
"I have always been told that I must look out for myself in New
York, and I begin to realize it. How does it happen you are so far
away from the hotel?"
"It is my afternoon off."
They sat and chatted of Colorado, about which Rupert felt
considerable curiosity. At the end of fifteen minutes their attention
was drawn to a man of prosperous appearance who seemed in
trouble. He paused as he reached their bench, and asked anxiously,
"Has either of you seen a young man, nicely dressed and carrying a
cane?" and he went on to describe Clarence Clayton.
"Yes," answered Packard and Rupert, simultaneously, "the fellow was
sitting here less than half an hour since."
"He has stolen my gold watch," said the new acquaintance.
"He tried to sell it to me. He said it cost a hundred and fifty dollars."
"So it did, and more, too."
"He offered it for fifty dollars."
"How did it happen that you did not buy it?"
"I was about to do so, but this boy told me he was a confidence
man."
"Then you knew him?" asked the stranger.
"Yes," answered Rupert.
"Can you suggest any way in which I can recover my property?"
"Yes, sir. Report the matter to the police, and ask to have Richard
Darke, a well-known detective, put on the case. I will give you a line
to him. He will know at once who it is."
"I will do so. Where can I find you again?"
"At the Somerset Hotel, on Broadway."
"Thank you. If I receive it I will gladly compensate you for your
suggestion."
"I thank you, but do not wish any compensation. If I can defeat this
man's dishonest scheme I shall feel well repaid."
"Our cunning friend will soon be overhauled, I suspect," said the
cattleman. "Did you say you were off this afternoon?"
"Yes, sir."
"I am new to the city and want a guide. Are you open to an
engagement?"
"Yes, sir," answered Rupert, with a smile. "But I don't care for pay."
"Then we don't go. Business is business, and there is no reason why
I should take up your time without paying you a fair sum."
"Just as you like, sir."
The two spent the next three or four hours in visiting different
objects of interest in New York. The Colorado man seemed much
pleased with his young companion.
"You must come out to Colorado some time, Rupert," he said. "You
are a boy who would succeed there, or indeed anywhere. We have
some men come out there who are failures at the East, and they are
surprised that they don't succeed in the West. But I tell you that it
takes as much brains to win success in Colorado as in New York."
"Is that always the case? I have heard of men getting rich in the
West who were poor at home."
"That is true. Perhaps they were in the wrong business. I don't mind
saying that was the case with me. I was in the insurance business in
Hartford, but I wasn't particularly well adapted for it. I couldn't talk.
Out in Colorado I have learned to understand cattle, and they have
made me rich."
"Mr. Clayton can talk."
"Yes, a little too well. Unfortunately he is not honest, and a
dishonest man ought not to thrive anywhere. In Colorado he
wouldn't live wrong. Thieves are summarily dealt with."
About seven o'clock Mr. Packard invited Rupert to dine with him at
Delmonico's.
Rupert had heard a great deal about this celebrated restaurant, and
was glad to accept the invitation.
 CHAPTER XVIII.
A SCENE AT DELMONICO'S.

The two friends entered Delmonico's on the Broadway side, and took
seats at one of the windows. Rupert, after giving the order, looked
about him. He was curious to see that famous restaurant. He was
destined to a surprise. At the second table, sitting with his back to
Mr. Packard and himself, was a person whom he had the best reason
to remember.
It was Clarence Clayton.
He touched Mr. Packard's arm, and silently pointed to Clayton.
"Well, I'll be jiggered!" exclaimed the cattleman, in surprise. "That
fellow has got nerve."
Mr. Clayton was evidently enjoying himself. Beside his plate stood a
pint bottle of champagne of Delmonico's special brand. His dinner
would probably involve an expense of five dollars.
"He must have sold or pawned the watch," suggested Rupert in a
low voice.
At this moment Clayton looked around. He at once recognized the
two whom he had last seen at the Battery.
"So we meet again?" he said, coolly.
"Yes," answered the cattleman. "You appear to be having a pleasant
time."
"I generally do," returned Clayton.
"You seem to have negotiated a loan."
"I met a party who seemed to know more about gold watches than
you do."
"I congratulate you," said Packard, dryly.
Clayton returned to the discussion of his dinner, and soon the two
friends were served.
"Shall I order some champagne, Rupert?" asked the man from
Colorado.
"Not for me. I have promised my mother to avoid drink."
"You are wise. Far be it from me to tempt you. I have seen too much
of the evil done by intemperance."
Clarence Clayton evidently had no such objection to drink. He
drained the bottle, and calling for a cigar, leaned back in his chair,
with a self-satisfied smile.
"That fellow is a curiosity," Packard said. "He probably has good
abilities, and would meet with success in an honest career. He has
made poor use of his talents. I wonder if he ever reflects upon the
inevitable end of his dishonesty?"
"It doesn't seem to trouble him much," returned Rupert.
Neither he nor Clayton observed the quiet entrance of a small,
unobtrusive man, with sharp eyes, who, taking rapid glances at the
guests, moved towards the table occupied by the adventurer.
Sitting in pleasant enjoyment of his cigar, Clayton's attention was
drawn by a slight tap on his shoulder. Looking up in momentary
impatience he saw the newcomer at his side.
Stifling an ejaculation he stared at him in dismay.
"Mr. Clayton," said Detective Darke, in a low voice, "I see you know
me."
"No, I can't say I have the pleasure," stammered Clayton.
"You are polite to call it a pleasure. I am Richard Darke."
"Can I offer you a glass of champagne, Mr. Darke?"
"There doesn't seem to be any left in the bottle."
"I will order another."
"I won't put you to that trouble. I have business with you, and must
request you to go with me."
"But——"
"I can take no denial," said the detective, sharply. "Go up to the
desk, settle your bill, and then we will go out together. There will be
no scene, and no one will know my errand, if you obey my
directions."
Clayton went up to the desk, paid his check, and then, turning to the
detective, said,
"I am at your service."
By this time Rupert noticed what was going on, and silently called
the attention of Mr. Packard to it.
"Poor chap!" said the cattleman, as Clayton and his unwelcome
companion left the restaurant, "his punishment has come sooner
than I anticipated. He will be punished, but I am afraid the owner of
the watch stands a poor chance of recovering his property."
"Probably he will get possession of the pawn ticket and so secure
the watch, though it may cost him twenty-five dollars."
"It will be some time before the thief gets another such dinner as he
has eaten to-night."
After supper Packard said,
"Are you feeling tired, Rupert?"
"Oh, no, sir."
"Then suppose we go to some theatre."
"All right, sir. What theatre do you prefer?"
"I leave the choice to you."
"Palmer's Theatre is very near."
"Then let us go there."
They reached the theatre just as the curtain was rising. Mr. Packard
bought two choice seats, and they were soon seated in the
orchestra. As soon as he had a chance to look about him, Rupert
discovered to his surprise that Mr. Lorimer and Julian were sitting
directly in front of him. At the sound of his voice Julian turned, and
was greatly surprised to see the bell-boy occupying as high-priced a
seat as himself. When the first act was ended, he took measures to
gratify his curiosity.
"I am surprised to see you here," he said.
"It is a mutual surprise," responded Rupert.
"You know what I mean. It is not usual to meet bell-boys in
orchestra seats."
"I was not asked at the ticket office what was my employment."
"Are you here alone?"
"No; let me introduce my friend, Mr. Packard, of Colorado."
Julian glanced at the cattleman, and was not impressed. Mr.
Packard's clothing was by no means stylish. Julian naturally
supposed him to be a person of small means and no particular
consequence. He gave him a slight nod, and turned his face towards
the stage.
"What is the name of that boy?" asked the cattleman.
"Julian Lorimer."
"Is he related to Stephen Lorimer?"
"Stephen Lorimer is his father. Do you know him?" asked Rupert, in
surprise.
"Stephen Lorimer is a cousin of mine."
"There he is, next to Julian."
Mr. Lorimer's ears caught the sound of his own name, and, turning,
he recognized Rupert, but not his cousin.
"You here, Rupert Rollins?" he said, in surprise.
"Yes, sir. Do you know this gentleman with me?"
Stephen Lorimer regarded the cattleman blankly.
"No," he answered slowly. "I don't know him."
"Perhaps you will remember the name of Giles Packard," said the
cattleman, but his tone was cold and not cordial.
"Are you Giles Packard?"
"Yes."
Stephen Lorimer looked embarrassed.
"I hope you are prosperous," he said.
"Thank you—I am doing well now."
"Where do you live?"
"In Colorado."
"Ah! Mines?"
"No, cattle."
"Call and see me. Rupert will tell you where I may be found."
"I may do so."
"Is he a cowboy?" asked Julian, in an audible whisper.
Giles Packard heard the words and he looked at Rupert with a smile.
"He is like his father," he said.
They did not again speak. After the play Stephen Lorimer went out
of the theatre without even a look at his new-found relative. Rupert
and the man from Colorado, following slowly, made the best of their
way down Broadway to the Somerset House.
"How came you to know Stephen Lorimer?" asked Packard.
"He and my father were in business together in Buffalo some years
since. They failed, and I have always believed that my father was
defrauded. At any rate he lost everything, while his old partner had
money enough to start in the dry-goods business in New York."
"History repeats itself," said Packard. "Many years ago, when I was
twenty-two, I was the partner of Stephen Lorimer."
"You!"
"Yes. In fact I furnished three-fourths of the capital. At the end of
eighteen months we failed. I never could understand why, for our
business had been good. Stephen kept the books, and I examined
without being able to understand them. The upshot of it was that I
was thrown upon the world penniless, while he soon went into
business for himself in another place. I have not seen Lorimer for
twenty years, till accident brought us together to-night."
"I am glad you are prosperous again."
"Yes. I have far more money than when I belonged to the firm of
Lorimer & Packard."
"Perhaps Mr. Lorimer would take you in as partner again."
"I have no desire to be associated with him in any way. I believe him
to be a thoroughly dishonest man. I am sorry that your father has
suffered also at his hands."
Rupert accompanied Mr. Packard to the hotel, having agreed to
relieve another bell-boy from midnight till six o'clock the next
morning.
When he reached the hotel he found it a scene of excitement. The
bell of No. 61 had been ringing violently for some time.
The other bell-boy had come downstairs in a panic.
"I can't get into No. 61," he reported. "There is somebody dead or
murdered there."
 CHAPTER XIX.
WHAT HAPPENED IN NO. 61.

"Come upstairs with me, Rupert," said Mr. Malcolm, the clerk.
"You've got a head on your shoulders. We'll soon find out what's the
matter."
They ascended in the elevator to the third floor, and made their way
hurriedly to No. 61.
There was a sound of a child crying inside. Mr. Malcolm tried the
door but it was locked.
"Open the door!" he called out.
"I can't," was the answer, in a young child's voice. "It's locked."
"Can't you turn the key?"
"No; I don't know how."
"You will have to get through the transom," said the clerk. "If we
only had a step-ladder."
"Lift me up and I'll get through," said Rupert. "I have practiced in a
gymnasium."
"Very well, if you think you can."
The clerk bent over, and Rupert, standing on his shoulders, was
lifted so that he could reach the transom.
Then, by a skillful movement, he raised himself still farther till he
could look inside.
"What do you see?" asked Malcolm.
"There is a man lying on his face on the floor. He must have had a fit
or something."
"Can you get through and lower yourself to the floor?"
"I think so. I will try."
"It is the only way to get into the room."
In very quick time Rupert accomplished his object. He turned the
key and opened the door.
It was as he had said. A man lay prone upon the floor, and beside
him, crying bitterly, was a pretty little boy of five, who was evidently
very much frightened.
"Papa sick," he said.
Malcolm bent over the prostrate man, and tearing open his vest
placed his hand on his heart.
"The man is dead!" he said, gravely, turning to Rupert.
The child was undressed, and the appearance of the bed showed
that he at least had occupied it.
"How long has your papa been lying here?" asked Malcolm.
"I don't know. I woke up a little while ago, and I saw him on the
floor."
"Is he cold?" asked Rupert.
"Yes; he must have been lying here for some time. Probably he was
about to undress, when he had an attack of some kind, and fell as
we see him. Call Dr. Bancroft."
A physician from Massachusetts was one of the guests of the hotel,
and occupied Room 57.
Summoned by Rupert, he entered the room, and immediately made
an examination of the body.
"Died of heart disease!" he said, briefly.
"Will papa soon be well?" asked the little boy, anxiously.
"We can tell better to-morrow," said the physician, pityingly. "You
had better go with this gentleman, so as not to disturb your father,
and we will do what we can for him."
Soothed by this assurance, for the little fellow did not understand
that his father was beyond earthly help, the boy was led away and
put in charge of a sympathetic lady guest for the night.
"Has he been dead long, doctor?" asked Malcolm.
"Probably for over an hour. What is his name?"
"I have forgotten. It is on the register."
"Perhaps we may find a letter in his pocket that will throw light on
the matter."
Malcolm put his hand in the inside coat pocket and drew out, first, a
letter addressed to
Paul Harvey,
Albany,
New York.
The other had no envelope and seemed to be an open letter. It ran
thus:

"To whom it may concern—

"My doctor tells me that I am liable at any moment to drop
dead from heart disease. I do not dread death for myself, but
when I think of my little Fred, soon to be left fatherless, as he is
already motherless, I am filled with anxiety. I am practically
 alone in the world, and there is no one to whom I can confide.
Should death come to me suddenly, I trust some kind-hearted
person will adopt Freddie, and supply a father's place to him. In
my inside vest pocket will be found securities amounting to
eleven hundred dollars. After defraying my funeral expenses
there will probably be a thousand dollars left. I leave it to any
one who will undertake the care and maintenance of my dear
little boy.
Paul Harvey."

The three looked at one another after the clerk had read the letter.
"Here is a responsibility for some one," said Dr. Bancroft. "I wish it
were in my power to take the little boy, but I am only here as a
guest, and circumstances will not permit."
"I am a bachelor, and should find it impossible to assume such a
charge," said the clerk, "though I feel for the little fellow."
An inspiration had come to Rupert. His heart had gone out to the
little boy so tragically deprived of his natural protector.
"I will take the little boy if you are willing," he said.
"You! A boy! What can you do with him?" asked Malcolm.
"I am boarding in a nice family," he said. "I will put him under the
care of Mrs. Benton, who has a young son of her own."
"But do you realize what a responsibility you are assuming?"
"I do, and I am not afraid. I never had a little fellow, and I shall be
very fond of Fred."
"What do you think, doctor?" asked the clerk.
"I think from the little I know of this boy, that, though a young
guardian, he will be a reliable one. I recommend that Fred, if that's
his name, be put under his charge."
"In that case, according to the father's direction, the money will go
to Rupert."
"Please take charge of it, Mr. Malcolm, till the funeral is over. Then
we will place it in some bank."
"It will not go very far towards paying for the boy's board and
education. He can't be more than five or six."
"When it is gone I will support him."
No objection was made, and it was agreed that Rupert should have
the custody of the little orphan, not yet conscious of his loss.
 CHAPTER XX.
MR. PACKARD'S GIFT.

It was not until the next day that Giles Packard knew of the tragedy
in No. 61. He had gone to bed at once on reaching the hotel, and
had not heard of Rupert's adopting a child.
"What is this I hear, Rupert?" he asked, on meeting the bell-boy. "I
hear you have an adopted son."
"Yes," answered Rupert, with a smile.
"Won't you get tired of the care and responsibility?"
"I think not."
"Besides, there will be considerable expense."
"The money left by his father will pay that till I am older and am
earning more."
"Not many boys of your age would dare to assume such a charge."
"Perhaps not, but Fred is such a sweet boy I cannot help loving
him."
"Look here, Rupert, won't you let me share the expense? I am rich
and have no family ties?"
"Thank you, Mr. Packard. I am very much obliged to you, but I
should like to feel that I am Fred's sole guardian. I want him to learn
to love me."
"I don't know but you are right. I won't interfere if you don't wish
me to."
That evening Rupert took Fred to Mr. Benton's.
"I have brought you another boarder," he said.
Mrs. Benton looked surprised.
"Is it a relation of yours?" she asked.
"He is my son."
The good lady looked amazed.
"My adopted son," amended Rupert, with a smile; and then he told
her of the sudden death at the hotel, and little Fred's bereavement.
Mrs. Benton's heart went out to the little orphan, and she stooped
and kissed him.
"Will you live with me?" she asked.
"I am going to live with him," said little Fred, taking Rupert's hand.
"He will live here, too."
"Then I will stay," answered the child, gravely. "I am to stay with
him till papa comes back."
They had told the little boy that his father had gone on a long
journey, and wished him to stay with Rupert during his absence. He
had acquiesced quietly, for he was a docile child, and transferred his
affection to Rupert, of whose love he felt assured.
"Now, Mrs. Benton, I must make a bargain with you for Fred's
board."
Mrs. Benton at first refused to accept anything, protesting that a
child would be little expense, but Rupert told her that the father had
left money, and finally induced her to accept three dollars a week.
"I am afraid that is too little," said the bell-boy.
"No; it will help pay the rent, and I shall like to have Freddie here as
a companion for Harry."
So it was arranged, and the little boy was provided with a happy and
comfortable home at small expense.
Two days later Giles Packard sought out Rupert during an interval of
the bell-boy's labors.
"How is the little boy?" asked the cattleman.
"He is well, and he seems to be happy. He thinks his father is away
on a journey."
"The journey we must all take some time," said Packard, gravely.
"Then you won't accept my help towards paying for the child's
maintenance?"
"It won't be necessary, Mr. Packard. I am to pay only three dollars a
week for his board."
"His clothing will cost something."
"Mrs. Benton will manage that. She says it won't cost over fifty
dollars a year."
"I foresaw that you wouldn't let me help support the boy, so I have
got even with you in another way."
"How is that?" asked Rupert, puzzled.
Mr. Packard, smiled.
"I decided to make you a present," he said. "You won't refuse that?"
"No; I am sure you are a good friend, and I won't reject your
kindness."
Rupert fancied Mr. Packard might be intending to give him fifty
dollars, or something like that, and he felt that it would be
ungracious to refuse.
The man from Colorado drew from his pocket a large-sized envelope,
and from it took a legal document.
"This," he said, "is a deed of two lots in Harlem, not far from One-
hundred-and-twenty-fifth Street. The deed is made out to you, and
establishes your ownership."
"I didn't know you had any lots in Harlem," said Rupert, in surprise.
"Neither had I till yesterday. I bought them through a real estate
agent on Third Avenue, after carefully considering several others."
"But, Mr. Packard, they must have cost you a good deal of money."
"Two thousand dollars."
"And you give me such a valuable present?"
"Yes, Rupert, and I am glad to do so. Don't think I have pinched
myself to do it. I am a rich man, and I haven't a chick or child,
except—well, except you," he continued, with a smile.
"I don't know that I ought to accept such a handsome present,
though I fully appreciate your great kindness."
"I don't quite see that you have any choice. The deed is made out in
your name, and in due time you will find that you will have to pay
taxes on them."
"Then I suppose I must submit. I don't know how to thank you."
"Then don't do it. It would make me feel awkward. I will give you
some good advice before I leave you. Those lots I believe will
advance in price very rapidly. Building is going on very near them,
and they are in the path of improvement. My advice is that you hold
on to them at least five years. They may realize you a small
fortune."
"I will certainly be guided by your advice. Do you know, Mr. Packard,
I imagine there are very few bell-boys in New York who are as rich
as I am?"
"I don't think I have ever heard of a bell-boy millionaire," said the
cattleman, smiling, "though I hope the one before me may make the
first exception to a general rule. Did I tell you that I expect to start
on my return to Colorado to-morrow?"
"So soon as that?"
"Yes; I have received news from my agent there—good news, mind
—that makes it advisable for me to abridge my visit. May I hope that
you will write me sometimes?"
"I shall be glad to do so, Mr. Packard."
"Mind, it is a compact. Some time I expect you to visit me out
there."
"When my child gets a little older," said Rupert, with a smile.
"And if at any time you find the expense too great for your means,
let me help you."
"I will."
So the two friends parted, and Rupert resumed his regular routine as
a bell-boy.
 CHAPTER XXI.
RUPERT BECOMES A CONFIDANT.

Some three months later Rupert's attention was called to a boy of

seventeen or thereabouts, with long black hair and a high forehead,
who registered as a guest, and took one of the cheapest rooms in
the hotel. The boy seemed to have no companion, and to know very
little about the city.
"Can you direct me to Palmer's Theatre?" he asked, rather
diffidently.
"It is on Broadway, corner of Thirtieth Street," answered Rupert.
"And Daly's?"
"That is nearly opposite, on the other side of Broadway."
The boy took out a memorandum-book and noted down these
addresses.
"What can he want at those theatres?" thought Rupert.
Of course he might want to buy a reserved seat in advance, but
Rupert did not think it likely.
After getting his information the boy went out (it was about ten
o'clock), and did not reappear till four o'clock in the afternoon.
Rupert noticed him as he entered the hotel, and observed that he
looked anxious and despondent. He did not go upstairs at once, but
sank into a chair near Rupert, and apparently gave way to sorrowful
reflections.
"He has some secret trouble," thought the bell-boy. "If he would
speak to me I might be able to comfort him."
On the impulse of the moment he went up to the young guest, and
asked, in a low tone of sympathy,
"Are you in any trouble?"
The boy started, flushed, and looked at Rupert half suspiciously. But
there was something so friendly and sympathetic in Rupert's face
that he was assured of his being a safe confidant.
"Yes," he said, "I am in trouble."
"If you will tell me, perhaps I can help you."
The boy looked about him hesitatingly.
"I shouldn't like to tell you here," he answered. "There are too many
people round."
"I shall be at leisure after six o'clock. Will that do?"
"Yes. Could you come up to my room?"
"I will come with pleasure."
"I want a confidant. I want advice. You are younger than I am—at
least you look so—but you have lived in the city while I am from the
country."
"At any rate I will give you the best advice I can."
"Thank you. I feel better for having found a friend. I will go and take
a walk, and you will find me here at six o'clock."
When Rupert got through work he found the boy waiting for him in
the same place.
"I can go upstairs with you now."
"All right!" said the young guest, rising from his seat quickly. "We will
take the elevator, for my room is on the top floor."
"In business hours," said Rupert, "I am not allowed to use the
elevator. Now I am no longer a bell-boy, but your visitor."
The room was a small hall bedroom. It was one that was let for
seventy-five cents a day, while the better and larger rooms ranged
upwards to a dollar and a half. The room contained one chair only.
"Please take a seat," said the young host.
"But where will you sit?"
"I will sit on the bed. I don't know but you will laugh at me," he
went on, "when I tell you what brought me to New York."
"Oh, no. I shall not laugh at you. But first, as we are to be friends,
let me tell you my name and ask yours. I am Rupert Rollins."
"That is a nice name. It sounds like a story name. Mine is Leslie
Waters."
"Where do you live?"'
"I was born and brought up in Rahway. That is in New Jersey, about
twenty miles from New York. My father lives about a mile from the
village. He has a small farm."
"And you were brought up to work on the farm?"
"Well, it isn't exactly a farm, but we raise vegetables and fruits for
the New York market. I went to school till a year ago. Then I
graduated, and since then I have worked for my father."
"Did you like it?"
"No, I don't like working on land. I feel," continued Leslie, flushing,
"that I was born for something better and nobler. Besides, I don't
want to live in the country. I prefer the city. There's something going
on here."
"Yes, that is true."
"And I wanted to be in the excitement. I'd rather live half as long in
the city. You can live more here in a year than in the country in two
years."
"Was there any particular thing that you wished to do?"
"Yes, I am coming to that. When I attended school there was one
exercise that many of the boys did not like, but I did. I liked to
declaim. I began with such pieces as 'Casabianca'—you know that,
don't you?"
"Oh, yes," said Rupert, smiling. "I have spoken it more than once
myself."
"But of course I got beyond that after a while. I used to speak
pieces from Shakespeare and other dramatic authors. There was one
I liked to speak in particular. It begins:

"The warrior bowed his crested head and tamed his heart of
fire,
And sued the haughty king to free his long-imprisoned sire."

"Yes, I know the poem."

"I got a prize for speaking it at one of our closing examinations,"
said Leslie, proudly. "Would you like to have me speak it for you
now?"
"I afraid it would attract attention in some of the neighboring rooms,
as it is a spirited piece."
Leslie looked disappointed but continued. "Then I have spoken
'Young Lochinvar' also—I liked that."
"Did you never speak any prose pieces?"
"No, I didn't care for prose. I like poetry best. I wish we were alone,
so I could speak something for you."