Introduction to Probability and

Statistics for Engineers and Scientists

6th Edition Sheldon M. Ross
Visit to download the full and correct content document:
https://textbookfull.com/product/introduction-to-probability-and-statistics-for-engineers
-and-scientists-6th-edition-sheldon-m-ross/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Probability statistics for engineers scientists Walpole

https://textbookfull.com/product/probability-statistics-for-
engineers-scientists-walpole/

Statistics for Engineers and Scientists William Navidi

https://textbookfull.com/product/statistics-for-engineers-and-
scientists-william-navidi/

Probability and statistics for computer scientists

Third Edition Michael Baron

https://textbookfull.com/product/probability-and-statistics-for-
computer-scientists-third-edition-michael-baron/

Statistics and Probability with Applications for

Engineers and Scientists Using MINITAB, R and JMP 2nd
Edition Bhisham C. Gupta

https://textbookfull.com/product/statistics-and-probability-with-
applications-for-engineers-and-scientists-using-minitab-r-and-
jmp-2nd-edition-bhisham-c-gupta/
Statistics for Engineers and Scientists 5th Edition
William Navidi

https://textbookfull.com/product/statistics-for-engineers-and-
scientists-5th-edition-william-navidi/

Miller Freund s probability and statistics for

engineers Freund

https://textbookfull.com/product/miller-freund-s-probability-and-
statistics-for-engineers-freund/

Probability, Statistics, and Stochastic Processes for

Engineers and Scientists (Mathematical Engineering,
Manufacturing, and Management Sciences) First Edition
Aliakbar Montazer Haghighi
https://textbookfull.com/product/probability-statistics-and-
stochastic-processes-for-engineers-and-scientists-mathematical-
engineering-manufacturing-and-management-sciences-first-edition-
aliakbar-montazer-haghighi/

Turbulence an introduction for scientists and engineers

Davidson

https://textbookfull.com/product/turbulence-an-introduction-for-
scientists-and-engineers-davidson/

Introduction to Probability and Statistics Metric

Edition 1925-2009) Mendenhall

https://textbookfull.com/product/introduction-to-probability-and-
statistics-metric-edition-1925-2009-mendenhall/
 Introduction to Probability and
Statistics for Engineers and Scientists
Introduction to Probability
and Statistics for
Engineers and Scientists
Sixth Edition

Sheldon M. Ross
University of Southern California
Los Angeles, CA, United States
Academic Press is an imprint of Elsevier
125 London Wall, London EC2Y 5AS, United Kingdom
525 B Street, Suite 1650, San Diego, CA 92101, United States
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
Copyright © 2021 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means, electronic
or mechanical, including photocopying, recording, or any information storage and retrieval system,
without permission in writing from the publisher. Details on how to seek permission, further
information about the Publisher’s permissions policies and our arrangements with organizations such
as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website:
www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the
Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience
broaden our understanding, changes in research methods, professional practices, or medical treatment
may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating
and using any information, methods, compounds, or experiments described herein. In using such
information or methods they should be mindful of their own safety and the safety of others, including
parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume
any liability for any injury and/or damage to persons or property as a matter of products liability,
negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas
contained in the material herein.

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library

ISBN: 978-0-12-824346-6

For information on all Academic Press publications

visit our website at https://www.elsevier.com/books-and-journals

Publisher: Katey Birtcher

Editorial Project Manager: Sara Valentino
Production Project Manager: Rukmani Krishnan
Designer: Patrick Ferguson
Typeset by VTeX
 For
Elise
 Contents

PREFACE ...........................................................................................................xiii
CHAPTER 1 Introduction to statistics............................................................ 1
1.1 Introduction ....................................................................... 1
1.2 Data collection and descriptive statistics........................ 1
1.3 Inferential statistics and probability models................... 2
1.4 Populations and samples ................................................. 3
1.5 A brief history of statistics................................................ 4
Problems ........................................................................... 7
CHAPTER 2 Descriptive statistics................................................................ 11
2.1 Introduction ..................................................................... 11
2.2 Describing data sets ....................................................... 12
2.2.1 Frequency tables and graphs ............................. 12
2.2.2 Relative frequency tables and graphs................ 14
2.2.3 Grouped data, histograms, ogives, and
stem and leaf plots.............................................. 16
2.3 Summarizing data sets................................................... 19
2.3.1 Sample mean, sample median, and sample
mode..................................................................... 19
2.3.2 Sample variance and sample standard
deviation ............................................................... 24
2.3.3 Sample percentiles and box plots ...................... 26
2.4 Chebyshev’s inequality ................................................... 29
2.5 Normal data sets ............................................................ 33
2.6 Paired data sets and the sample
correlation coefficient..................................................... 36
2.7 The Lorenz curve and Gini index.................................... 43
2.8 Using R ............................................................................ 48
Problems ......................................................................... 52
CHAPTER 3 Elements of probability............................................................ 63
3.1 Introduction ..................................................................... 63
3.2 Sample space and events............................................... 64 vii
viii Contents

3.3 Venn diagrams and the algebra of events..................... 66

3.4 Axioms of probability ...................................................... 67
3.5 Sample spaces having equally likely outcomes............ 70
3.6 Conditional probability.................................................... 75
3.7 Bayes’ formula ................................................................ 79
3.8 Independent events ........................................................ 86
Problems ......................................................................... 89
CHAPTER 4 Random variables and expectation......................................... 99
4.1 Random variables ........................................................... 99
4.2 Types of random variables ........................................... 102
4.3 Jointly distributed random variables........................... 105
4.3.1 Independent random variables......................... 111
4.3.2 Conditional distributions................................... 114
4.4 Expectation.................................................................... 117
4.5 Properties of the expected value ................................. 121
4.5.1 Expected value of sums of random variables.. 124
4.6 Variance ......................................................................... 128
4.7 Covariance and variance of sums of
random variables .......................................................... 132
4.8 Moment generating functions...................................... 138
4.9 Chebyshev’s inequality and the weak law of large
numbers ........................................................................ 139
Problems ....................................................................... 142
CHAPTER 5 Special random variables ...................................................... 151
5.1 The Bernoulli and binomial random variables ........... 151
5.1.1 Using R to calculate binomial probabilities..... 157
5.2 The Poisson random variable....................................... 158
5.2.1 Using R to calculate Poisson probabilities ...... 166
5.3 The hypergeometric random variable ......................... 167
5.4 The uniform random variable ...................................... 171
5.5 Normal random variables ............................................ 179
5.6 Exponential random variables ..................................... 190
5.6.1 The Poisson process ......................................... 193
5.6.2 The Pareto distribution ..................................... 196
5.7 The gamma distribution ............................................... 199
5.8 Distributions arising from the normal ........................ 201
5.8.1 The chi-square distribution .............................. 201
5.8.2 The t -distribution .............................................. 206
5.8.3 The F -distribution ............................................. 208
5.9 The logistics distribution .............................................. 209
5.10 Distributions in R .......................................................... 210
Problems ....................................................................... 212
 Contents ix

CHAPTER 6 Distributions of sampling statistics ...................................... 221

6.1 Introduction ................................................................... 221
6.2 The sample mean ......................................................... 222
6.3 The central limit theorem ............................................ 224
6.3.1 Approximate distribution of the sample mean 227
6.3.2 How large a sample is needed?........................ 230
6.4 The sample variance..................................................... 230
6.5 Sampling distributions from a normal population ..... 231
6.5.1 Distribution of the sample mean...................... 232
6.5.2 Joint distribution of X and S 2 ........................... 232
6.6 Sampling from a finite population ............................... 234
Problems ....................................................................... 238
CHAPTER 7 Parameter estimation............................................................ 245
7.1 Introduction ................................................................... 245
7.2 Maximum likelihood estimators .................................. 246
7.2.1 Estimating life distributions ............................. 255
7.3 Interval estimates ......................................................... 257
7.3.1 Confidence interval for a normal mean when
the variance is unknown ................................... 262
7.3.2 Prediction intervals ........................................... 268
7.3.3 Confidence intervals for the variance of a
normal distribution ........................................... 269
7.4 Estimating the difference in means of two normal
populations.................................................................... 270
7.5 Approximate confidence interval for the mean of a
Bernoulli random variable ........................................... 275
7.6 Confidence interval of the mean of the exponential
distribution ................................................................... 280
7.7 Evaluating a point estimator ........................................ 281
7.8 The Bayes estimator..................................................... 287
Problems ....................................................................... 292
CHAPTER 8 Hypothesis testing ................................................................. 305
8.1 Introduction ................................................................... 305
8.2 Significance levels ........................................................ 306
8.3 Tests concerning the mean of a normal population... 307
8.3.1 Case of known variance .................................... 307
8.3.2 Case of unknown variance: the t -test.............. 319
8.4 Testing the equality of means of two normal
populations.................................................................... 326
8.4.1 Case of known variances .................................. 326
8.4.2 Case of unknown variances .............................. 328
8.4.3 Case of unknown and unequal variances ........ 333
8.4.4 The paired t -test................................................ 333
8.5 Hypothesis tests concerning the variance of a
normal population ........................................................ 336
x Contents

8.5.1 Testing for the equality of variances of two

normal populations ........................................... 337
8.6 Hypothesis tests in Bernoulli populations .................. 339
8.6.1 Testing the equality of parameters in two
Bernoulli populations........................................ 342
8.7 Tests concerning the mean of a Poisson distribution 345
8.7.1 Testing the relationship between two
Poisson parameters .......................................... 346
Problems ....................................................................... 348
CHAPTER 9 Regression.............................................................................. 365
9.1 Introduction ................................................................... 365
9.2 Least squares estimators of the
regression parameters................................................. 367
9.3 Distribution of the estimators ...................................... 371
9.4 Statistical inferences about the
regression parameters................................................. 377
9.4.1 Inferences concerning β ................................... 377
9.4.2 Inferences concerning α ................................... 386
9.4.3 Inferences concerning the mean response
α + βx0 ............................................................... 386
9.4.4 Prediction interval of a future response .......... 389
9.4.5 Summary of distributional results ................... 392
9.5 The coefficient of determination and the sample
correlation coefficient................................................... 392
9.6 Analysis of residuals: assessing the model ................ 395
9.7 Transforming to linearity.............................................. 396
9.8 Weighted least squares ................................................ 400
9.9 Polynomial regression.................................................. 406
9.10 Multiple linear regression ............................................ 410
9.10.1 Predicting future responses ............................. 420
9.10.2 Dummy variables for categorical data ............. 424
9.11 Logistic regression models for binary output data..... 425
Problems ....................................................................... 429
CHAPTER 10 Analysis of variance ............................................................... 453
10.1 Introduction ................................................................... 453
10.2 An overview ................................................................... 454
10.3 One-way analysis of variance....................................... 456
10.3.1 Using R to do the computations ....................... 463
10.3.2 Multiple comparisons of sample means.......... 466
10.3.3 One-way analysis of variance with unequal
sample sizes ...................................................... 468
10.4 Two-factor analysis of variance: introduction and
parameter estimation................................................... 470
10.5 Two-factor analysis of variance: testing hypotheses.. 474
 Contents xi

10.6 Two-way analysis of variance with interaction ........... 479

Problems ....................................................................... 487
CHAPTER 11 Goodness of fit tests and categorical data analysis............. 499
11.1 Introduction ................................................................... 499
11.2 Goodness of fit tests when all parameters are
specified ........................................................................ 500
11.2.1 Determining the critical region by simulation. 506
11.3 Goodness of fit tests when some parameters are
unspecified .................................................................... 508
11.4 Tests of independence in contingency tables.............. 510
11.5 Tests of independence in contingency tables having
fixed marginal totals..................................................... 514
11.6 The Kolmogorov–Smirnov goodness of fit test for
continuous data............................................................. 517
Problems ....................................................................... 522
CHAPTER 12 Nonparametric hypothesis tests........................................... 529
12.1 Introduction ................................................................... 529
12.2 The sign test.................................................................. 529
12.3 The signed rank test ..................................................... 533
12.4 The two-sample problem ............................................. 538
12.4.1 Testing the equality of multiple probability
distributions....................................................... 541
12.5 The runs test for randomness ..................................... 544
Problems ....................................................................... 547
CHAPTER 13 Quality control ........................................................................ 555
13.1 Introduction ................................................................... 555
13.2 Control charts for average values: the x control
chart............................................................................... 556
13.2.1 Case of unknown μ and σ ................................. 559
13.3 S-control charts ............................................................ 564
13.4 Control charts for the fraction defective ..................... 567
13.5 Control charts for number of defects.......................... 569
13.6 Other control charts for detecting changes in the
population mean ........................................................... 573
13.6.1 Moving-average control charts ........................ 573
13.6.2 Exponentially weighted moving-average
control charts .................................................... 576
13.6.3 Cumulative sum control charts ........................ 581
Problems ....................................................................... 583
CHAPTER 14 Life testing∗ ............................................................................ 591
14.1 Introduction ................................................................... 591
14.2 Hazard rate functions ................................................... 591

∗ Optional chapter.
xii Contents

14.3 The exponential distribution in life testing.................. 594

14.3.1 Simultaneous testing — stopping at the rth
failure ................................................................. 594
14.3.2 Sequential testing ............................................. 599
14.3.3 Simultaneous testing — stopping by a fixed
time .................................................................... 603
14.3.4 The Bayesian approach ..................................... 606
14.4 A two-sample problem ................................................. 607
14.5 The Weibull distribution in life testing......................... 609
14.5.1 Parameter estimation by least squares........... 611
Problems ....................................................................... 613
CHAPTER 15 Simulation, bootstrap statistical methods, and
permutation tests................................................................... 619
15.1 Introduction ................................................................... 619
15.2 Random numbers ......................................................... 619
15.2.1 The Monte Carlo simulation approach............. 622
15.3 The bootstrap method .................................................. 623
15.4 Permutation tests ......................................................... 631
15.4.1 Normal approximations in permutation tests . 634
15.4.2 Two-sample permutation tests ........................ 637
15.5 Generating discrete random variables........................ 639
15.6 Generating continuous random variables................... 641
15.6.1 Generating a normal random variable............. 643
15.7 Determining the number of simulation runs
in a Monte Carlo study.................................................. 644
Problems ....................................................................... 645
CHAPTER 16 Machine learning and big data.............................................. 649
16.1 Introduction ................................................................... 649
16.2 Late flight probabilities ................................................ 650
16.3 The naive Bayes approach............................................ 651
16.3.1 A variation of naive Bayes approach ................ 654
16.4 Distance-based estimators. The k-nearest
neighbors rule............................................................... 657
16.4.1 A distance-weighted method............................ 658
16.4.2 Component-weighted distances....................... 659
16.5 Assessing the approaches............................................ 660
16.6 When characterizing vectors are quantitative ............ 662
16.6.1 Nearest neighbor rules ..................................... 662
16.6.2 Logistics regression .......................................... 663
16.7 Choosing the best probability: a bandit problem........ 664
Problems ....................................................................... 666
APPENDIX OF TABLES ................................................................................... 669
INDEX .............................................................................................................. 673
Preface

The sixth edition of this book continues to demonstrate how to apply probabil-
ity theory to gain insight into real, everyday statistical problems and situations.
As in the previous editions, carefully developed coverage of probability mo-
tivates probabilistic models of real phenomena and the statistical procedures
that follow. This approach ultimately results in an intuitive understanding of
statistical procedures and strategies most often used by practicing engineers
and scientists.
This book has been written for an introductory course in statistics or in proba-
bility and statistics for students in engineering, computer science, mathematics,
statistics, and the natural sciences. As such it assumes knowledge of elementary
calculus.

Organization and coverage

Chapter 1 presents a brief introduction to statistics, presenting its two branches
of descriptive and inferential statistics, and a short history of the subject and
some of the people whose early work provided a foundation for work done
today.
The subject matter of descriptive statistics is then considered in Chapter 2.
Graphs and tables that describe a data set are presented in this chapter, as are
quantities that are used to summarize certain of the key properties of the data
set.
To be able to draw conclusions from data, it is necessary to have an under-
standing of the data’s origination. For instance, it is often assumed that the
data constitute a “random sample” from some population. To understand ex-
actly what this means and what its consequences are for relating properties of
the sample data to properties of the entire population, it is necessary to have
some understanding of probability, and that is the subject of Chapter 3. This
chapter introduces the idea of a probability experiment, explains the concept
of the probability of an event, and presents the axioms of probability. xiii
xiv Preface

Our study of probability is continued in Chapter 4, which deals with the

important concepts of random variables and expectation, and in Chapter 5,
which considers some special types of random variables that often occur in ap-
plications. Such random variables as the binomial, Poisson, hypergeometric,
normal, uniform, gamma, chi-square, t, and F are presented.
In Chapter 6, we study the probability distribution of such sampling statistics
as the sample mean and the sample variance. We show how to use a remark-
able theoretical result of probability, known as the central limit theorem, to
approximate the probability distribution of the sample mean. In addition, we
present the joint probability distribution of the sample mean and the sample
variance in the important special case in which the underlying data come from
a normally distributed population.
Chapter 7 shows how to use data to estimate parameters of interest. For in-
stance, a scientist might be interested in determining the proportion of Mid-
western lakes that are afflicted by acid rain. Two types of estimators are studied.
The first of these estimates the quantity of interest with a single number (for
instance, it might estimate that 47 percent of Midwestern lakes suffer from acid
rain), whereas the second provides an estimate in the form of an interval of
values (for instance, it might estimate that between 45 and 49 percent of lakes
suffer from acid rain). These latter estimators also tell us the “level of con-
fidence” we can have in their validity. Thus, for instance, whereas we can be
pretty certain that the exact percentage of afflicted lakes is not 47, it might very
well be that we can be, say, 95 percent confident that the actual percentage is
between 45 and 49.
Chapter 8 introduces the important topic of statistical hypothesis testing,
which is concerned with using data to test the plausibility of a specified hy-
pothesis. For instance, such a test might reject the hypothesis that fewer than
44 percent of Midwestern lakes are afflicted by acid rain. The concept of the
p-value, which measures the degree of plausibility of the hypothesis after the
data have been observed, is introduced. A variety of hypothesis tests concerning
the parameters of both one and two normal populations are considered. Hy-
pothesis tests concerning Bernoulli and Poisson parameters are also presented.
Chapter 9 deals with the important topic of regression. Both simple linear
regression — including such subtopics as regression to the mean, residual anal-
ysis, and weighted least squares — and multiple linear regression are consid-
ered.
Chapter 10 introduces the analysis of variance. Both one-way and two-way
(with and without the possibility of interaction) problems are considered.
Chapter 11 is concerned with goodness of fit tests, which can be used to test
whether a proposed model is consistent with data. In it we present the classical
chi-square goodness of fit test and apply it to test for independence in con-
tingency tables. The final section of this chapter introduces the Kolmogorov–
 Preface xv

Smirnov procedure for testing whether data come from a specified continuous
probability distribution.
Chapter 12 deals with nonparametric hypothesis tests, which can be used
when one is unable to suppose that the underlying distribution has some spec-
ified parametric form (such as normal).
Chapter 13 considers the subject matter of quality control, a key statistical tech-
nique in manufacturing and production processes. A variety of control charts,
including not only the Shewhart control charts but also more sophisticated
ones based on moving averages and cumulative sums, are considered.
Chapter 14 deals with problems related to life testing. In this chapter, the ex-
ponential, rather than the normal, distribution plays the key role.
In Chapter 15, we consider the statistical inference techniques of bootstrap sta-
tistical methods and permutation tests. We first show how probabilities can be
obtained by simulation and then how to utilize simulation in these statistical
inference approaches.
Chapter 16, new to this edition, introduces machine learning and big data
techniques. These are methods that are applicable in situations where one has
a large amount of data that can be used to estimate probabilities without as-
suming any particular probability model. For instance, we consider situations
where one wants to estimate the probability that an experiment, characterized
by a vector (x1 , . . . , xn ), will be a success. When the characterizing vectors are
qualitative in nature, such techniques as the naive Bayes approach, and nearest
neighbor rules are studied. In cases where the components of the characteristic
vector are quantitative, we also study logistic regression models.
Aside from the newly added Chapter 16, the most important change in this
edition is the use of the statistical software R. No previous experience with R is
necessary, and we incorporate its use throught the text. Aside from additional
subsections devoted to R. we also have the newly added Section 2.7, dealing
with Lorenz curves and the Gini index. There are also many new examples and
problems in this edition. In addition, the sixth edition contains a multitude of
small changes designed to even further increase the clarity of the text’s presen-
tations and arguments.

Supplemental materials

Solutions manual for instructors is available at: https://textbooks.elsevier.com/

web/Manuals.aspx?isbn=9780128243466.
xvi Preface

Acknowledgments
We thank the following people for their helpful comments on material of the
sixth edition:

■ Gideon Weiss, University of Haifa

■ N. Balakrishnan, McMaster University
■ Mark Brown, Columbia University
■ Rohitha Goonatilake, Texas A and M University
■ Steve From, University of Nebraska at Omaha
■ Subhash Kochar, Portland State University
■ Sumona Mondal, Mathematics, Clarkson University
■ Kamel Belbahri, Mathematics and Statistics, Université de Montréal
■ Anil Aswani, Industrial Engineering and Operations Research, University
of California, Berkeley

as well as all those reviewers who asked to remain anonymous.

 CHAPTER 1

Introduction to statistics

1.1 Introduction
It has become accepted in today’s world that in order to learn about something,
you must first collect data. Statistics is the art of learning from data. It is con-
cerned with the collection of data, its subsequent description, and its analysis,
which often leads to the drawing of conclusions.

1.2 Data collection and descriptive statistics

Sometimes a statistical analysis begins with a given set of data: For instance, the
government regularly collects and publicizes data concerning yearly precipita-
tion totals, earthquake occurrences, the unemployment rate, the gross domestic
product, and the rate of inflation. Statistics can be used to describe, summarize,
and analyze these data.
In other situations, data are not yet available; in such cases statistical theory can
be used to design an appropriate experiment to generate data. The experiment
chosen should depend on the use that one wants to make of the data. For
instance, suppose that an instructor is interested in determining which of two
different methods for teaching computer programming to beginners is most
effective. To study this question, the instructor might divide the students into
two groups, and use a different teaching method for each group. At the end
of the class the students can be tested and the scores of the members of the
different groups compared. If the data, consisting of the test scores of members
of each group, are significantly higher in one of the groups, then it might seem
reasonable to suppose that the teaching method used for that group is superior.
It is important to note, however, that in order to be able to draw a valid con-
clusion from the data, it is essential that the students were divided into groups
in such a manner that neither group was more likely to have the students with
greater natural aptitude for programming. For instance, the instructor should
not have let the male class members be one group and the females the other.
For if so, then even if the women scored significantly higher than the men, it
would not be clear whether this was due to the method used to teach them,
or to the fact that women may be inherently better than men at learning pro- 1

Introduction to Probability and Statistics for Engineers and Scientists. https://doi.org/10.1016/B978-0-12-824346-6.00010-7

Copyright © 2021 Elsevier Inc. All rights reserved.
2 CHAP TER 1: Introduction to statistics

gramming skills. The accepted way of avoiding this pitfall is to divide the class
members into the two groups “at random.” This term means that the division
is done in such a manner that all possible choices of the members of a group
are equally likely.
At the end of the experiment, the data should be described. For instance, the
scores of the two groups should be presented. In addition, summary mea-
sures such as the average score of members of each of the groups should be
presented. This part of statistics, concerned with the description and summa-
rization of data, is called descriptive statistics.

1.3 Inferential statistics and probability models

After the preceding experiment is completed and the data are described and
summarized, we hope to be able to draw a conclusion about which teaching
method is superior. This part of statistics, concerned with the drawing of con-
clusions, is called inferential statistics.
To be able to draw a conclusion from the data, we must take into account the
possibility of chance. For instance, suppose that the average score of members
of the first group is quite a bit higher than that of the second. Can we conclude
that this increase is due to the teaching method used? Or is it possible that the
teaching method was not responsible for the increased scores but rather that
the higher scores of the first group were just a chance occurrence? For instance,
the fact that a coin comes up heads 7 times in 10 flips does not necessarily
mean that the coin is more likely to come up heads than tails in future flips.
Indeed, it could be a perfectly ordinary coin that, by chance, just happened to
land heads 7 times out of the total of 10 flips. (On the other hand, if the coin
had landed heads 47 times out of 50 flips, then we would be quite certain that
it was not an ordinary coin.)
To be able to draw logical conclusions from data, we usually make some as-
sumptions about the chances (or probabilities) of obtaining the different data
values. The totality of these assumptions is referred to as a probability model for
the data.
Sometimes the nature of the data suggests the form of the probability model
that is assumed. For instance, suppose that an engineer wants to find out what
proportion of computer chips, produced by a new method, will be defective.
The engineer might select a group of these chips, with the resulting data being
the number of defective chips in this group. Provided that the chips selected
were “randomly” chosen, it is reasonable to suppose that each one of them is
defective with probability p, where p is the unknown proportion of all the chips
produced by the new method that will be defective. The resulting data can then
be used to make inferences about p.
 1.4 Populations and samples 3

In other situations, the appropriate probability model for a given data set will
not be readily apparent. However, careful description and presentation of the
data sometimes enable us to infer a reasonable model, which we can then try
to verify with the use of additional data.
Because the basis of statistical inference is the formulation of a probability
model to describe the data, an understanding of statistical inference requires
some knowledge of the theory of probability. In other words, statistical infer-
ence starts with the assumption that important aspects of the phenomenon
under study can be described in terms of probabilities; it then draws conclu-
sions by using data to make inferences about these probabilities.

1.4 Populations and samples

In statistics, we are interested in obtaining information about a total collection
of elements, which we will refer to as the population. The population is often too
large for us to examine each of its members. For instance, we might have all the
residents of a given state, or all the television sets produced in the last year by
a particular manufacturer, or all the households in a given community. In such
cases, we try to learn about the population by choosing and then examining a
subgroup of its elements. This subgroup of a population is called a sample.
If the sample is to be informative about the total population, it must be, in
some sense, representative of that population. For instance, suppose that we are
interested in learning about the age distribution of people residing in a given
city, and we obtain the ages of the first 100 people to enter the town library. If
the average age of these 100 people is 46.2 years, are we justified in concluding
that this is approximately the average age of the entire population? Probably
not, for we could certainly argue that the sample chosen in this case is probably
not representative of the total population because usually more young students
and senior citizens use the library than do working-age citizens.
In certain situations, such as the library illustration, we are presented with a
sample and must then decide whether this sample is reasonably representative
of the entire population. In practice, a given sample generally cannot be as-
sumed to be representative of a population unless that sample has been chosen
in a random manner. This is because any specific nonrandom rule for selecting
a sample often results in one that is inherently biased toward some data values
as opposed to others.
Thus, although it may seem paradoxical, we are most likely to obtain a repre-
sentative sample by choosing its members in a totally random fashion without
any prior considerations of the elements that will be chosen. In other words, we
need not attempt to deliberately choose the sample so that it contains, for in-
stance, the same gender percentage and the same percentage of people in each
profession as found in the general population. Rather, we should just leave it
4 CHAP TE R 1: Introduction to statistics

up to “chance” to obtain roughly the correct percentages. Once a random sam-

ple is chosen, we can use statistical inference to draw conclusions about the
entire population by studying the elements of the sample.

1.5 A brief history of statistics

A systematic collection of data on the population and the economy was begun
in the Italian city-states of Venice and Florence during the Renaissance. The
term statistics, derived from the word state, was used to refer to a collection of
facts of interest to the state. The idea of collecting data spread from Italy to the
other countries of Western Europe. Indeed, by the first half of the 16th cen-
tury it was common for European governments to require parishes to register
births, marriages, and deaths. Because of poor public health conditions this
last statistic was of particular interest.
The high mortality rate in Europe before the 19th century was due mainly to
epidemic diseases, wars, and famines. Among epidemics, the worst were the
plagues. Starting with the Black Plague in 1348, plagues recurred frequently
for nearly 400 years. In 1562, as a way to alert the King’s court to consider
moving to the countryside, the City of London began to publish weekly bills of
mortality. Initially these mortality bills listed the places of death and whether
a death had resulted from plague. Beginning in 1625 the bills were expanded
to include all causes of death.
In 1662 the English tradesman John Graunt published a book entitled Natural
and Political Observations Made upon the Bills of Mortality. Table 1.1, which notes
the total number of deaths in England and the number due to the plague for
five different plague years, is taken from this book.

Table 1.1 Total Deaths in England.

Year Burials Plague Deaths
1592 25,886 11,503
1593 17,844 10,662
1603 37,294 30,561
1625 51,758 35,417
1636 23,359 10,400
Source: John Graunt, Observations Made upon the Bills of Mortal-
ity. 3rd ed. London: John Martyn and James Allestry (1st ed. 1662).

Graunt used London bills of mortality to estimate the city’s population. For in-
stance, to estimate the population of London in 1660, Graunt surveyed house-
holds in certain London parishes (or neighborhoods) and discovered that, on
average, there were approximately 3 deaths for every 88 people. Dividing by 3
shows that, on average, there was roughly 1 death for every 88/3 people. Be-
cause the London bills cited 13,200 deaths in London for that year, Graunt
 1.5 A brief history of statistics 5

Table 1.2 John Graunt’s Mortality Table.

Age at Death Number of Deaths per
100 Births
0–6 36
6–16 24
16–26 15
26–36 9
36–46 6
46–56 4
56–66 3
66–76 2
76 and greater 1
Note: The categories go up to but do not include the right-hand
value. For instance, 0–6 means all ages from 0 up through 5.

estimated the London population to be about

13,200 × 88/3 = 387,200

Graunt used this estimate to project a figure for all England. In his book he
noted that these figures would be of interest to the rulers of the country, as
indicators of both the number of men who could be drafted into an army and
the number who could be taxed.
Graunt also used the London bills of mortality — and some intelligent guess-
work as to what diseases killed whom and at what age — to infer ages at death.
(Recall that the bills of mortality listed only causes and places at death, not the
ages of those dying.) Graunt then used this information to compute tables giv-
ing the proportion of the population that dies at various ages. Table 1.2 is one
of Graunt’s mortality tables. It states, for instance, that of 100 births, 36 people
will die before reaching age 6, 24 will die between the age of 6 and 15, and so
on.
Graunt’s estimates of the ages at which people were dying were of great interest
to those in the business of selling annuities. Annuities are the opposite of life
insurance in that one pays in a lump sum as an investment and then receives
regular payments for as long as one lives.
Graunt’s work on mortality tables inspired further work by Edmund Halley
in 1693. Halley, the discoverer of the comet bearing his name (and also the
man who was most responsible, by both his encouragement and his financial
support, for the publication of Isaac Newton’s famous Principia Mathematica),
used tables of mortality to compute the odds that a person of any age would
live to any other particular age. Halley was influential in convincing the insurers
of the time that an annual life insurance premium should depend on the age
of the person being insured.
6 CHAP TE R 1: Introduction to statistics

Following Graunt and Halley, the collection of data steadily increased through-
out the remainder of the 17th and on into the 18th century. For instance, the
city of Paris began collecting bills of mortality in 1667, and by 1730 it had
become common practice throughout Europe to record ages at death.
The term statistics, which was used until the 18th century as a shorthand for the
descriptive science of states, became in the 19th century increasingly identified
with numbers. By the 1830s the term was almost universally regarded in Britain
and France as being synonymous with the “numerical science” of society. This
change in meaning was caused by the large availability of census records and
other tabulations that began to be systematically collected and published by
the governments of Western Europe and the United States beginning around
1800.
Throughout the 19th century, although probability theory had been developed
by such mathematicians as Jacob Bernoulli, Karl Friedrich Gauss, and Pierre-
Simon Laplace, its use in studying statistical findings was almost nonexistent,
because most social statisticians at the time were content to let the data speak
for themselves. In particular, statisticians of that time were not interested in
drawing inferences about individuals, but rather were concerned with the soci-
ety as a whole. Thus, they were not concerned with sampling but rather tried
to obtain censuses of the entire population. As a result, probabilistic infer-
ence from samples to a population was almost unknown in 19th century social
statistics.
It was not until the late 1800s that statistics became concerned with inferring
conclusions from numerical data. The movement began with Francis Galton’s
work on analyzing hereditary genius through the uses of what we would now
call regression and correlation analysis (see Chapter 9), and obtained much
of its impetus from the work of Karl Pearson. Pearson, who developed the
chi-square goodness of fit tests (see Chapter 11), was the first director of the
Galton Laboratory, endowed by Francis Galton in 1904. There Pearson origi-
nated a research program aimed at developing new methods of using statistics
in inference. His laboratory invited advanced students from science and in-
dustry to learn statistical methods that could then be applied in their fields.
One of his earliest visiting researchers was W.S. Gosset, a chemist by training,
who showed his devotion to Pearson by publishing his own works under the
name “Student.” (A famous story has it that Gosset was afraid to publish un-
der his own name for fear that his employers, the Guinness brewery, would be
unhappy to discover that one of its chemists was doing research in statistics.)
Gosset is famous for his development of the t-test (see Chapter 8).
Two of the most important areas of applied statistics in the early 20th century
were population biology and agriculture. This was due to the interest of Pear-
son and others at his laboratory and also to the remarkable accomplishments
of the English scientist Ronald A. Fisher. The theory of inference developed by
these pioneers, including among others Karl Pearson’s son Egon and the Pol-
 Problems 7

Table 1.3 The Changing Definition of Statistics.

Statistics has then for its object that of presenting a faithful representation of a state at a
determined epoch. (Quetelet, 1849)
Statistics are the only tools by which an opening can be cut through the formidable thicket
of difficulties that bars the path of those who pursue the Science of man. (Galton, 1889)
Statistics may be regarded (i) as the study of populations, (ii) as the study of variation, and
(iii) as the study of methods of the reduction of data. (Fisher, 1925)
Statistics is a scientific discipline concerned with collection, analysis, and interpretation of
data obtained from observation or experiment. The subject has a coherent structure based
on the theory of Probability and includes many different procedures which contribute to
research and development throughout the whole of Science and Technology. (E. Pearson,
1936)
Statistics is the name for that science and art which deals with uncertain inferences —
which uses numbers to find out something about nature and experience. (Weaver, 1952)
Statistics has become known in the 20th century as the mathematical tool for analyzing
experimental and observational data. (Porter, 1986)
Statistics is the art of learning from data. (this book, 2020)

ish born mathematical statistician Jerzy Neyman, was general enough to deal
with a wide range of quantitative and practical problems. As a result, after the
early years of the 20th century a rapidly increasing number of people in sci-
ence, business, and government began to regard statistics as a tool that was
able to provide quantitative solutions to scientific and practical problems (see
Table 1.3).
Nowadays the ideas of statistics are everywhere. Descriptive statistics are fea-
tured in every newspaper and magazine. Statistical inference has become indis-
pensable to public health and medical research, to engineering and scientific
studies, to marketing and quality control, to education, to accounting, to eco-
nomics, to meteorological forecasting, to polling and surveys, to sports, to
insurance, to gambling, and to all research that makes any claim to being sci-
entific. Statistics has indeed become ingrained in our intellectual heritage.

Problems
1. An election will be held next week and, by polling a sample of the voting
population, we are trying to predict whether the Republican or Demo-
cratic candidate will prevail. Which of the following methods of selection
is likely to yield a representative sample?
a. Poll all people of voting age attending a college basketball game.
b. Poll all people of voting age leaving a fancy midtown restaurant.
c. Obtain a copy of the voter registration list, randomly choose 100
names, and question them.
8 CHAP TER 1: Introduction to statistics

d. Use the results of a television call-in poll, in which the station asked
its listeners to call in and name their choice.
e. Choose names from the telephone directory and call these people.
2. The approach used in Problem 1(e) led to a disastrous prediction in the
1936 presidential election, in which Franklin Roosevelt defeated Alfred
Landon by a landslide. A Landon victory had been predicted by the Lit-
erary Digest. The magazine based its prediction on the preferences of a
sample of voters chosen from lists of automobile and telephone owners.
a. Why do you think the Literary Digest’s prediction was so far off?
b. Has anything changed between 1936 and now that would make you
believe that the approach used by the Literary Digest would work
better today?
3. A researcher is trying to discover the average age at death for people in the
United States today. To obtain data, the obituary columns of the New York
Times are read for 30 days, and the ages at death of people in the United
States are noted. Do you think this approach will lead to a representative
sample?
4. To determine the proportion of people in your town who are smokers, it
has been decided to poll people at one of the following local spots:
a. the pool hall;
b. the bowling alley;
c. the shopping mall;
d. the library.
Which of these potential polling places would most likely result in a rea-
sonable approximation to the desired proportion? Why?
5. A university plans on conducting a survey of its recent graduates to deter-
mine information on their yearly salaries. It randomly selected 200 recent
graduates and sent them questionnaires dealing with their present jobs.
Of these 200, however, only 86 were returned. Suppose that the average
of the yearly salaries reported was $75,000.
a. Would the university be correct in thinking that $75,000 was a good
approximation to the average salary level of all of its graduates? Ex-
plain the reasoning behind your answer.
b. If your answer to part (a) is no, can you think of any set of conditions
relating to the group that returned questionnaires for which it would
be a good approximation?
6. An article reported that a survey of clothing worn by pedestrians killed
at night in traffic accidents revealed that about 80 percent of the victims
were wearing dark-colored clothing and 20 percent were wearing light-
colored clothing. The conclusion drawn in the article was that it is safer
to wear light-colored clothing at night.
a. Is this conclusion justified? Explain.
b. If your answer to part (a) is no, what other information would be
needed before a final conclusion could be drawn?
 Problems 9

7. Critique Graunt’s method for estimating the population of London.

What implicit assumption is he making?
8. The London bills of mortality listed 12,246 deaths in 1658. Supposing
that a survey of London parishes showed that roughly 2 percent of the
population died that year, use Graunt’s method to estimate London’s
population in 1658.
9. Suppose you were a seller of annuities in 1662 when Graunt’s book was
published. Explain how you would make use of his data on the ages at
which people were dying.
10. Based on Graunt’s mortality table:
a. What proportion of people survived to age 6?
b. What proportion survived to age 46?
c. What proportion died between the ages of 6 and 36?
 CHAPTER 2

Descriptive statistics

2.1 Introduction

In this chapter we introduce the subject matter of descriptive statistics, and in

doing so learn ways to describe and summarize a set of data. Section 2.2 deals
with ways of describing a data set. Subsections 2.2.1 and 2.2.2 indicate how
data that take on only a relatively few distinct values can be described by using
frequency tables or graphs, whereas Subsection 2.2.3 deals with data whose set
of values is grouped into different intervals. Section 2.3 discusses ways of sum-
marizing data sets by use of statistics, which are numerical quantities whose
values are determined by the data. Subsection 2.3.1 considers three statistics
that are used to indicate the “center” of the data set: the sample mean, the
sample median, and the sample mode. Subsection 2.3.2 introduces the sam-
ple variance and its square root, called the sample standard deviation. These
statistics are used to indicate the spread of the values in the data set. Subsection
2.3.3 deals with sample percentiles, which are statistics that tell us, for instance,
which data value is greater than 95 percent of all the data. In Section 2.4 we
present Chebyshev’s inequality for sample data. This famous inequality gives
an upper bound to the proportion of the data that can differ from the sample
mean by more than k times the sample standard deviation. Whereas Cheby-
shev’s inequality holds for all data sets, we can in certain situations, which are
discussed in Section 2.5, obtain more precise estimates of the proportion of
the data that is within k sample standard deviations of the sample mean. In
Section 2.5 we note that when a graph of the data follows a bell-shaped form
the data set is said to be approximately normal, and more precise estimates are
given by the so-called empirical rule. Section 2.6 is concerned with situations
in which the data consist of paired values. A graphical technique, called the
scatter diagram, for presenting such data is introduced, as is the sample corre-
lation coefficient, a statistic that indicates the degree to which a large value of
the first member of the pair tends to go along with a large value of the second.
Section 2.7 is concerned with the income distribution of a population. It intro-
duces the Lorenz curve L(p), which gives the proportion of the total income of
a population that is earned by the lower 100p percent of wage earners. Also in-
troduced is the Gini index, which is a measure of the inequality of the income
distribution. Section 2.8 shows how to use R to analyze data sets. 11

Introduction to Probability and Statistics for Engineers and Scientists. https://doi.org/10.1016/B978-0-12-824346-6.00011-9

Copyright © 2021 Elsevier Inc. All rights reserved.
12 CHAPTER 2: Descriptive statistics

2.2 Describing data sets

The numerical findings of a study should be presented clearly, concisely, and in
such a manner that an observer can quickly obtain a feel for the essential char-
acteristics of the data. Over the years it has been found that tables and graphs
are particularly useful ways of presenting data, often revealing important fea-
tures such as the range, the degree of concentration, and the symmetry of the
data. In this section we present some common graphical and tabular ways for
presenting data.

2.2.1 Frequency tables and graphs

A data set having a relatively small number of distinct values can be conve-
niently presented in a frequency table. For instance, Table 2.1 is a frequency table
for a data set consisting of the starting yearly salaries (to the nearest thousand
dollars) of 42 recently graduated students with B.S. degrees in electrical engi-
neering. Table 2.1 tells us, among other things, that the lowest starting salary
of $57,000 was received by four of the graduates, whereas the highest salary
of $70,000 was received by a single student. The most common starting salary
was $62,000, and was received by 10 of the students.

Table 2.1 Starting Yearly Salaries.

Starting Salary Frequency
57 4
58 1
59 3
60 5
61 8
62 10
63 0
64 5
66 2
67 3
70 1

Data from a frequency table can be graphically represented by a line graph that
plots the distinct data values on the horizontal axis and indicates their fre-
quencies by the heights of vertical lines. A line graph of the data presented in
Table 2.1 is shown in Figure 2.1.
When the lines in a line graph are given added thickness, the graph is called a
bar graph. Figure 2.2 presents a bar graph.
Another type of graph used to represent a frequency table is the frequency poly-
gon, which plots the frequencies of the different data values on the vertical axis,
 2.2 Describing data sets 13

FIGURE 2.1
Starting salary data.

FIGURE 2.2
Bar graph for starting salary data.
14 CHAPTER 2: Descriptive statistics

FIGURE 2.3
Frequency polygon for starting salary data.

and then connects the plotted points with straight lines. Figure 2.3 presents a
frequency polygon for the data of Table 2.1.

2.2.2 Relative frequency tables and graphs

Consider a data set consisting of n values. If f is the frequency of a particular
value, then the ratio f/n is called its relative frequency. That is, the relative fre-
quency of a data value is the proportion of the data that have that value. The
relative frequencies can be represented graphically by a relative frequency line
or bar graph or by a relative frequency polygon. Indeed, these relative frequency
graphs will look like the corresponding graphs of the absolute frequencies ex-
cept that the labels on the vertical axis are now the old labels (that gave the
frequencies) divided by the total number of data points.
Example 2.2.a. Table 2.2 is a relative frequency table for the data of Table 2.1.
The relative frequencies are obtained by dividing the corresponding frequencies
of Table 2.1 by 42, the size of the data set. 

A pie chart is often used to indicate relative frequencies when the data are not
numerical in nature. A circle is constructed and then sliced into different sec-
Another random document with
no related content on Scribd:
The Project Gutenberg eBook of Sous les déodars
This ebook is for the use of anyone anywhere in the United States
and most other parts of the world at no cost and with almost no
restrictions whatsoever. You may copy it, give it away or re-use it
under the terms of the Project Gutenberg License included with this
ebook or online at www.gutenberg.org. If you are not located in the
United States, you will have to check the laws of the country where
you are located before using this eBook.

Title: Sous les déodars

Author: Rudyard Kipling

Translator: Albert Savine

Release date: September 27, 2023 [eBook #71738]

Language: French

Original publication: Paris: Stock, 1910

Credits: Véronique Le Bris, Laurent Vogel, Chuck Greif and the

Online Distributed Proofreading Team at
https://www.pgdp.net (This file was produced from images
generously made available by the Bibliothèque nationale de
France (BnF/Gallica))

*** START OF THE PROJECT GUTENBERG EBOOK SOUS LES

DÉODARS ***
 SOUS LES DÉODARS
L’auteur et l’éditeur déclarent réserver leurs droits de traduction et de reproduction pour tous les
pays, y compris la Suède et la Norvège.
Ce volume a été déposé au Ministère de l’Intérieur (section de la librairie) en septembre 1909.

TABLE DES MATIÈRES

DU MÊME AUTEUR ET DU MÊME TRADUCTEUR

Simples Contes des Collines.

Nouveaux Contes des Collines.
Trois Troupiers.
Autres Troupiers.
Au blanc et noir.

DU MÊME TRADUCTEUR

Juan Valera.—Le Commandeur Mendoza.

Narcis Oller.—Le Papillon, préface d’Émile Zola.
— Le Rapiat.
Jacinto Verdaguer.—L’Atlantide.
Emilia Pardo Bazan.—Le Naturalisme.
Henryk Sienkiewicz.—Pages d’Amérique.
Andrew Carnegie.—La Grande-Bretagne jugée par un Américain.
Elisabeth Barrett Browning.—Poèmes et poésies.
Th. de Quincey.—Souvenirs autobiographiques du Mangeur d’opium.
Th. Roosevelt.—La Vie au Rancho.
— Chasses et parties de chasse.
— La Conquête de l’Ouest.
— New-York.
Percy Bysshe Shelley.—Œuvres en prose.
Robert-L. Stevenson.—Enlevé!
Algernon C. Swinburne.—Nouveaux Poèmes et Ballades.
Oscar Wilde.—Le Crime de lord Arthur Savile.
— Le Portrait de monsieur W. H.
— Poèmes.
— Le Prêtre et l’Acolyte.
— Théâtre.—I.—Drames.
— Théâtre.—II.—Comédies.
A. Conan Doyle.—Mystères et Aventures.
— Le Parasite.
— La Grande Ombre.
— Un Début en médecine.
— Idylle de banlieue.
— Nouveaux Mystères et Aventures.

EN PRÉPARATION

Rudyard Kipling.—La Cité de l’épouvantable nuit.

— Lettres de marque.
— Au hasard de la vie.
A. Conan Doyle.—Rodney Stone.
— La merveilleuse Découverte de Raffles Haw.
Henryk Sienkiewicz.—La Préférée.
Armando Palacio Valdes.—L’Idylle d’un malade.
José-Maria de Pereda.—Au premier Vol.
Robert-L. Stevenson.—Les joyeux Drilles.
Bret Harte.—Maruja.

BIBLIOTHÈQUE COSMOPOLITE—Nº 38

RUDYARD KIPLING

Sous les Déodars

 Traduction d’ALBERT SAVINE
Et comme il est incapable
d’employer, d’utiliser
convenablement le court laps de
temps qui
lui fut confié en dépôt et qu’il le
gaspille d’une
façon ennuyeuse et morne en peines
et sots
tourments, en querelles, en plaisirs,
naturellement,
il réclame à grands cris l’héritage de
l’éternel avenir, pour que son mérite
puisse se
donner libre carrière,—ce qui
évidemment
est de toute justice.
(La Cité de l’épouvantable nuit.)

PARIS—Iᵉʳ
P.-V. STOCK, ÉDITEUR
155, RUE SAINT-HONORÉ, 155
DEVANT LE THÉATRE-FRANÇAIS
—
1910

De cet ouvrage il a été tiré à part,

sur papier de Hollande, huit exemplaires
numérotés et paraphés par l’éditeur.

A NONCE CASANOVA
20 Avril 1909.
 Nous connaissons tous le cèdre du Liban. Les Poètes de la Bible l’ont
chanté comme le plus altier et le plus superbe des arbres de l’Asie, et
Jussieu, au siècle le moins créateur de légendes, a renouvelé la sienne en le
rapportant à travers les flots de la Méditerranée et en le nourrissant de
partie de sa ration d’eau.
Nous en savons bien moins long sur le Déodar ou Déodara, le cèdre de
l’Himalaya. Bien qu’on l’ait acclimaté dans la forêt de Fontainebleau, ce
conifère aux rameaux flexibles et inclinés, à la feuille glauque et
blanchâtre, nous est à peu près inconnu.
Voici Kipling, qui va nous initier aux charmes des Déodars qui couvrent
les pentes de l’Himalaya.
N’était-il pas naturel qu’il empruntât, lors de ses débuts, le titre d’un de
ses recueils de nouvelles à un arbre aussi abondant au Jakko et aux bords
de Simla.
C’est, en effet, à la période de sa vie ou il venait d’écrire les Simples
Contes des Collines, Trois Troupiers et Au Blanc et Noir que se
rattachent les pages que nous présentons aujourd’hui au public français.
Nos lecteurs y retrouveront Madame Hauksbee en compagnie de son
amie Madame Mallowe, le Tertium quid dont Kipling leur a ailleurs
promis l’histoire, la Colline de l’Illusion. Ce sont d’anciennes
connaissances qu’on revoit volontiers.
A. S.
 L’ÉDUCATION D’OTIS YEERE
Dans les retraites du charmant
verger, «que Dieu bénisse tout nos
profits», disons-nous. Mais «que Dieu
bénisse nos pertes», voilà un souhait
qui convient mieux à notre situation.
(Le Berceau de verdure perdu[A].)

I
Cette histoire est celle d’un insuccès, mais la femme qui échoua disait
qu’on en pourrait faire un récit instructif et qui mériterait d’être imprimé
pour le plus grand profit de la génération nouvelle.
La génération nouvelle ne demande point à recevoir des leçons, étant
tout à fait prête à en donner à quiconque voudra bien lui en demander.
Qu’importe! Voici l’histoire.
Elle commence où doit commencer une histoire qui se respecte, c’est-à-
dire à Simla: c’est là que toutes commencent et que quelques-unes finissent
d’une façon funeste.
La méprise vint de ce qu’une femme des plus intelligentes commit une
maladresse, et ne la répara point.
Les hommes ont le droit reconnu de faire des faux pas; mais qu’une
femme intelligente commette une erreur, c’est en dehors des voies
régulières de la Nature et de la Providence.
Tous les braves gens savent en effet qu’une femme est la seule chose
infaillible qu’il y ait au monde, excepté le titre d’emprunt émis par le
gouvernement en 1879, et portant intérêt à quatre et demi pour cent.
Toutefois nous devons nous rappeler que six jours consécutifs passés à
répéter le rôle principal de l’Ange Déchu au Nouveau Théâtre de la Gaîté,
où les plâtres ne sont pas encore secs, c’était bien suffisant pour produire
une certaine rupture d’équilibre intellectuel, capable à son tour de conduire
à des excentricités.
Mistress Hauksbee arriva à la «fonderie» pour déjeuner avec mistress
Mallowe, son unique amie intime, car elle n’était en aucune façon femme à
frayer avec son sexe.
Et ce fut un déjeuner entre femmes, porté interdite à tout le monde.
Et toutes deux se mirent à parler chiffons, ce qui en français est
équivalent de «mystères.»
—J’ai joui d’une période de santé parfaite, dit mistress Hauksbee, le
déjeuner fini, et quand les deux dames furent confortablement installées
dans le petit boudoir qui communiquait avec la chambre à coucher de
mistress Mallowe:
—Ma chère petite, qu’est-ce qu’il a fait? dit avec douceur mistress
Mallowe.
Il est à remarquer que les dames d’un certain âge se traitent
mutuellement de «ma chère petite» tout comme des fonctionnaires qui ont
vingt-huit ans de service se disent: «Mon garçon,» entre employés de même
grade dans l’Annuaire.
—Il n’y a point de il dans l’affaire. Qui suis-je donc pour qu’on
m’impute toujours gratuitement quelque conquête imaginaire? Suis-je un
apache?
—Non, ma chère, mais il y a presque toujours un scalp en train de sécher
à l’entrée de votre wigwam, et un scalp tout frais.
C’était une allusion au petit Hawley qui avait pris l’habitude de courir
tout Simla à cheval, à la saison des pluies, pour aller rendre visite à mistress
Hauksbee.
Cette dame se mit à rire.
—Pour mes péchés, l’aide-major de Tyrconnel m’a condamnée, l’autre
soir, à me placer auprès du Mussuck. Chut! Ne riez pas. C’est un de mes
admirateurs les plus dévoués. Quand on servit les entremets—il faudrait
réellement que quelqu’un aille leur apprendre à faire les puddings, à
Tyrconnel,—le Mussuck fut enfin libre de se consacrer à mon service.
—La bonne âme! Je connais son appétit, dit mistress Mallowe. Est-ce
qu’il s’est mis, oh! est-ce qu’il s’est mis à faire sa cour?
 —Grâce à une faveur spéciale de la Providence, non. Il a expliqué
l’importance qu’il avait comme une des colonnes de l’Empire. Je n’ai point
ri.
—Lucy, je ne vous crois pas.
—Demandez au capitaine Sangar. Il était en face de nous. Je disais donc
que le Mussuck poitrinait.
—Il me semble que je le vois faisant la roue, dit d’un air pensif mistress
Mallowe, en grattant les oreilles de son fox-terrier.
—Je fus impressionnée comme il convenait, tout à fait comme il
convenait. Je bâillai franchement.
—Une surveillance sans trêve et l’art de jouer des uns contre les autres,
disait le Mussuck en engloutissant sa glace par pelletées, je vous en
réponds, mistress Hauksbee, voilà le secret de notre gouvernement.
Mistress Mallowe rit longtemps et gaîment:
—Et qu’avez-vous dit?
—M’avez-vous jamais vue embarrassée pour répondre? J’ai dit:
—C’est bien ce que j’ai remarqué dans mes relations avec vous.
Le Mussuck se gonfla d’orgueil.
Il va venir me voir demain. Le petit Hawley doit venir aussi.
—«Surveillance constante et l’art de jouer de l’un contre l’autre. Voilà,
mistress Hauksbee, voilà le secret de notre gouvernement». Et j’irai jusqu’à
dire que si nous pouvions pénétrer jusqu’au cœur du Mussuck, nous
verrions qu’il se regarde comme un homme de génie.
—Comme il est des deux autres choses. Il me plaît le Mussuck, et je ne
vous permettrai pas de lui donner des noms d’oiseau. Il m’amuse.
—Il vous a convertie vous aussi, à ce qu’il paraît. Parlez-moi de cette
période de santé parfaite et, je vous en prie, donnez à Tim une tape sur le
nez avec le coupe-papier. Ce chien aime trop le sucre. Prenez-vous du lait
dans votre thé?
—Non, merci. Polly, je suis lasse de cette vie: elle est vide.
—Mettez-vous à la dévotion dans ce cas. J’ai toujours dit que vous
finiriez par Rome.
—Cela se réduirait à planter là une demi-douzaine d’attachés en
uniforme rouge pour un seul costume noir, et si je jeûnais, il me viendrait
des rides, qui ne s’en iraient jamais, jamais. Avez-vous remarqué, ma chère,
que je vieillis!
—Merci de cette courtoisie, mais je vais vous la rendre. Oui, nous ne
sommes plus tout à fait, ni vous ni moi... comment dirai-je?
—Ce que nous avons été. «Je sens ça dans mes os,» pour parler comme
mistress Crossley. Polly, j’ai gâché ma vie.
—Comment ça?
—Le comment importe peu; mais je le sens. Je prétends devenir une
Puissance, avant de mourir.
—Alors soyez une Puissance. Vous avez de l’esprit assez pour faire
n’importe quoi... et la beauté.
Mistress Hauksbee brandit une cuiller à thé dans la direction de son
hôtesse.
—Polly, si vous m’accablez ainsi sous les compliments, j’en viendrai à
ne plus croire que vous êtes femme. Dites-moi comment faire pour devenir
une Puissance?
—Apprenez au Mussuck qu’il est le plus enchanteur et le plus svelte des
hommes d’Asie, il vous dira tout ce qui vous plaira, en gros et en détail.
—Fi du Mussuck! Je vise à devenir une Puissance intellectuelle, et non
une force motrice. Polly, je vais organiser un salon.
Mistress Mallowe se tourna languissamment sur le canapé et posa sa tête
sur sa main.
—Écoutez les paroles du Prophète, le fils de Baruch, dit-elle.
—Vous déciderez-vous à parler raisonnablement?
—C’est mon intention, ma chère, car je vois que vous êtes sur le point de
commettre une sottise.
—Je n’ai jamais de ma vie commis de sottise,—du moins de sottise pour
laquelle je n’aie pu trouver une explication, après coup.
—Sur le point de commettre une sottise, reprit mistress Mallowe sans se
déconcerter. A Simla, impossible d’organiser un salon. Un bar offrirait plus
de chances de succès.
—Peut-être. Mais pourquoi? Cela semble si facile.
—C’est justement en cela que la chose est difficile. Combien y a-t-il de
femmes intelligentes à Simla?
 —Deux: vous et moi, dit mistress Hauksbee sans l’ombre d’une
hésitation.
—Quelle modestie. Mistress Feardon vous en saurait gré. Et combien
d’hommes intelligents?
—Oh! une... des centaines, dit mistress Hauksbee, d’un air vague.
—Voilà l’erreur fatale! Il n’y en a pas un seul. Ils sont tous engagés
d’avance par le gouvernement. Voyez mon mari, par exemple. Jack a été un
homme intelligent. Je le dis: d’autres le diraient aussi. Le gouvernement lui
a mis le grappin dessus. Toutes ses idées, tous ses talents de causeur,—et
jadis il était vraiment un causeur de talent, même aux yeux de sa femme—
tout cela lui a été ôté par ce... cet évier de gouvernement. Il en est de même
pour tous les hommes qui ont quelque emploi ici. Je ne suppose pas qu’un
condamné russe sous le régime du knout soit fort propre à amuser le reste
de son équipe, et tout notre monde masculin est une troupe de forçats en
habits à dorures.
—Mais il y a des douzaines de...
—Je sais ce que vous allez dire: des douzaines de gens en congé, de gens
désœuvrés. Je l’admets, mais ils se répartissent en deux catégories
détestables: le civil qui serait enchanté de posséder la connaissance du
monde et la distinction du militaire, et le militaire qui serait adorable s’il
avait la culture du civil.
—Mot détestable. Les civils ont-ils de la culture? Je n’ai jamais étudié
cette espèce à fond.
—Ne vous gaussez pas de l’emploi de Jack. Oui: ils sont comme les
théières du bazar de Lakka, bonne matière, mais sans aucun chic. Ils n’en
peuvent mais, les pauvres mignons. Un civil ne commence à devenir
supportable qu’après avoir roulé par le monde une quinzaine d’années.
—Et un militaire?
—Quand il a servi pendant le même temps. Les jeunes de chaque
catégorie sont affreux. Vous en auriez par douzaines dans votre salon.
—Je ne le souffrirais pas, dit mistress Hauksbee avec une résolution
farouche, je dirais au portier de les balayer. Je mettrais leurs colonels et
leurs commissaires de planton à la porte pour les empêcher d’entrer. Je les
donnerais à la petite Topsham pour en faire joujou.
 —La petite Topsham vous saurait gré de ce cadeau. Mais revenons au
salon. Admettons que vous ayez réuni tous les hommes et toutes les femmes
ensemble, qu’en ferez-vous? Les faire causer? Mais ils se mettraient à
flirter d’un commun accord. Notre salon deviendrait un Peliti de bon ton, un
Hôtel de la Médisance, éclairé par des lampes.
—Il y a une certaine dose de raison dans cette remarque.
—Il y a toute la sagesse de ce monde. Certes, douze saisons passées à
Simla auraient dû vous apprendre qu’il est impossible de concentrer quoi
que ce soit dans l’Inde, et un salon ne peut réussir qu’à la condition d’être
permanent. En deux saisons, tout votre personnel serait dispersé d’un bout à
l’autre de l’Asie. Nous ne sommes guère que de petites boules de terre sur
les flancs des collines, et qu’un jour ou l’autre, la vallée aspirera de son
souffle. Nous avons perdu l’art de causer—du moins nos hommes l’ont
perdu.—Nous n’avons point de cohésion...
—George Eliot ressuscitée! interrompit malignement mistress Hauksbee.
—Et puis, ma chère railleuse, ni hommes, ni femmes n’ont
collectivement d’influence. Venez à la vérandah et jetons un coup d’œil sur
le Mail.
Les deux dames vinrent considérer la route qui se peuplait rapidement,
car tout Simla était dehors pour profiter d’un entracte entre averse et
brouillard.
—Que comptez-vous faire pour fixer ce flot? Regardez: voici le
Mussuck, un homme, qui est la bonté même. C’est une puissance dans le
pays, bien qu’il mange autant qu’un marchand des quatre saisons. Voici le
colonel Blone, le général Grucher, sir Dugald Delane et sir Henry
Haughton, et Mr. Jellalatty, tous des chefs de service, tous des gens
puissants.
—Et tous mes fervents admirateurs, dit mistress Hauksbee avec onction.
Sir Henry Haugton est fou de moi. Mais continuez.
—Pris à part, chacun d’eux est un homme de mérite. Réunis, ils ne sont
plus qu’une cohue d’Anglo-Indiens?. Qui s’intéresse à des propos d’Anglo-
Indiens? Votre Salon n’arriverait pas à souder ensemble les différents
ministères et à vous rendre maîtresse de l’Inde, ma chère. Tous ces gens-là
se mettraient à parler de leur boutique administrative et le feraient, en se
groupant dans votre salon, tant ils ont peur que leurs propos ne soient
surpris par les gens de condition inférieure. Ils ont oublié tout ce qu’ils ont
pu savoir de littérature et d’art... Quant aux femmes...
—La seule chose dont elles puissent causer, ce sont les dernières
Courses, ou les gaffes de leur dernière bonne. Ce matin, j’étais en visite
chez mistress Derwills...
—Vous croyez cela? Elles savent causer avec les petits officiers et les
petits officiers savent causer avec elles. Votre salon ferait admirablement
leur affaire, si vous respectiez les préjugés religieux du pays, et que vous
vous teniez amplement pourvue de Kala juggahs[B].
—Quantité de Kala juggahs! Oh! ma pauvre petite idée! Des Kala
juggahs dans un salon politique! Mais qui donc vous en a appris aussi long?
—C’est peut-être que j’en ai essayé moi-même ou bien que je connais
une femme qui en a essayé. J’ai fait un sermon en règle pour peser le pour
et le contre. La conclusion, c’est...
—Inutile d’achever... c’est le mot: néant! Polly, je vous remercie. Ces
maudites bêtes...
Et mistress Hauksbee, de la vérandah, montra de la main deux hommes
fendant la foule qui passaient au-dessous, et qui la saluèrent d’un coup de
chapeau.
—Ces mauvaises bêtes n’auront pas la joie de posséder un second hôtel
des Potins, ou un Peliti d’extra. Je renonce à l’idée de tenir un salon. Cela
me paraissait pourtant bien séduisant. Mais que faire? Il faut pourtant que je
fasse quelque chose.
—Pourquoi? N’y a-t-il pas Abana et Pharpar?
—Jack vous a rendue presque aussi malicieuse que lui. Il me faut cela,
naturellement. Je me lasse de tout et de tous, depuis une partie de campagne
au clair de lune, à Seepee, jusqu’aux charmes du Mussuck.
—Oui, ces choses-là arrivent tôt ou tard. Avez-vous encore assez de
vigueur pour tendre votre arc?
Mistress Hauksbee ferma la bouche d’un air rageur.
Puis elle se mit à rire.
—Je crois m’y voir. De grandes affiches rouges sur le Mail: «Mistress
Hauksbee! Irrévocablement: sa dernière représentation sur quelque scène
que ce soit. Qu’on se le dise!» Plus de danses, plus de promenades à cheval,
plus de petits déjeuners, plus de représentations théâtrales suivies de
soupers, plus de querelles à l’ami le plus aimé, le plus cher, plus d’escrime
avec un partenaire mal choisi qui n’a pas assez d’esprit pour habiller d’un
langage décent ce qu’il lui plaît d’appeler ses sentiments, plus d’exhibition
publique du Mussuck pendant que mistress Tarkass va, de maison en
maison, partout Simla, colporter d’horribles histoires sur mon compte! Plus
aucune de ces choses si profondément assommantes, abominables,
détestables, mais qui, tout de même, donnent tout son intérêt à l’existence!
Oui, je vois tout! Ne m’interrompez pas, Polly, je suis inspirée. Un «nuage»
à raies mauve et blanc sur mes superbes épaules, une place au cinquième
rang à la Gaîté, et les deux chevaux vendus! Vision délicieuse. Un fauteuil
confortable, où aboutissent trois courants d’air différents, dans chaque salle
de bal, et de beaux souliers amples, raisonnables, qui permettent à tous les
couples de trébucher en se rendant à la vérandah. Puis on va souper.
Pouvez-vous vous imaginer la scène? La cohue gloutonne est partie. Un
petit sous-lieutenant qui se fait prier, aussi rouge par tout son visage qu’un
baby auquel on vient de mettre de la poudre... On ferait vraiment bien de
tanner les petits sous-lieutenants avant de les exporter... Polly... La
maîtresse de maison le renvoyant à son service, il traverse la pièce d’un pas
furtif, dans ma direction, en tourmentant un gant deux fois trop grand pour
lui,—je déteste les gens qui portent les gants à la façon d’un pardessus,—et
tâche d’avoir l’air d’avoir pensé à cela pour la première fois: «Puis-je
havoir le plaisir de vous offrir mon bras pour le souper?» Alors, je me lève
avec le sourire que donne l’appétit. Tenez, comme ceci.
—Lucy, comment pouvez-vous être aussi absurde?
—Et je m’avance majestueusement à son bras. Comme cela! Après le
souper, je partirais de bonne heure, vous savez, parce que je craindrais de
m’enrhumer. Personne pour s’occuper de mon rickshaw, le mien, s’il vous
plaît. Je resterais là, avec ce «nuage» mauve et blanc sur la tête, pendant
que l’humidité trempe mes chers, mes vieux, mes respectables pieds, et que
Tom appelle à force de jurons et de cris l’équipage de la memsahib. Puis, on
rentre. On se couche à onze heures et demie. Voilà une vie vraiment
excellente, où l’on est réconfortée par les visites du Padri, qui vient à
l’instant même de conduire quelqu’un en terre quelque part là-bas.
Elle montra dans le lointain les pins qui cachaient le cimetière et reprit
avec un geste violemment dramatique:
 —Écoutez, je vois tout... tout jusqu’aux corsets! Quels corsets! Six
roupies huit aunas la paire, Polly, avec de la flanelle rouge, ou bien de la
lisière, n’est-ce pas? Ce qu’on met au bout de ces choses terribles! Je
pourrais vous en faire un dessin.
—Lucy, au nom du Ciel, finissez donc d’agiter les bras de cette façon
idiote. Songez qu’on peut vous voir de tout le Mail.
—Eh bien, qu’on voie! On croira que je m’exerce pour l’Ange Déchu.
Tenez, voici le Mussuck. Comme il se tient mal à cheval! Voyez.
Elle envoya, avec une grâce infinie, un baiser au vénérable
administrateur indien.
—A présent, voilà qui lui vaudra d’être blagué au Club, en ces termes
délicats qu’affectent ces brutes d’hommes, et le petit Hawley me rapportera
tout, en atténuant les détails de peur de me choquer. Ce garçon est trop bon
pour vivre longtemps, Polly. Je songe sérieusement à lui recommander de
donner sa démission et d’entrer dans le clergé. Dans l’état d’esprit où il se
trouve présentement, il m’obéirait. Heureux, heureux enfant!
—Jamais, dit mistress Mallowe avec une indignation affectée, jamais
vous ne déjeunerez plus ici, Lucinde, votre conduite est scandaleuse.
—C’est votre faute, répliqua mistress Hauksbee, pourquoi avoir voulu
me suggérer d’abdiquer? Rien que cela! Non, jamais! ja-a-mais! Je jouerai,
je danserai, je chevaucherai, je flirterai, je ferai des cancans, je dînerai en
ville, je m’approprierai les prisonniers légitimes de toutes les femmes qu’il
me plaira, jusqu’à ce que je tombe, ou qu’une femme plus forte que moi me
confonde devant tout Simla, et ma bouche ne sera plus que poussière et
cendres avant que je capitule ainsi.
Elle se dirigea vers le salon.
Mistress Mallowe la suivit et lui passa le bras autour de la taille.
—Il n’y a rien à redire à ma conduite, reprit mistress Hauksbee d’un air
de défi, et cherchant son mouchoir. Voilà dix soirs que je dîne en ville et
que je passe l’après-midi à répéter. Vous en seriez fatiguée vous-même. Je
suis seulement fatiguée, rien que fatiguée.
Mistress Mallowe ne témoigna point de compassion à mistress Hauksbee
et ne l’engagea point à aller se coucher. Elle lui donna une autre tasse de thé
et renoua la conversation.
—J’ai passé par là, moi aussi, ma chère, dit-elle.
 —Je m’en souviens, dit mistress Hauksbee, avec un rayonnement de
malice sur les traits, en 84, n’est-ce pas? La saison suivante, vous vous êtes
beaucoup moins surmenée.
Mistress Mallowe sourit d’un air de supériorité, d’un air de sphinx.
—Je suis devenue une Influence.
—Grands Dieux! mon enfant, vous ne vous êtes pas envolée parmi les
Théosophistes, et vous n’avez pas baisé le gros orteil de Bouddha, n’est-ce
pas? J’ai voulu m’affilier jadis, mais on m’a écartée comme sceptique—ce
qui m’ôte toute chance de perfectionner ma pauvre petite intelligence.
—Non, je n’ai pas théosophisé. Jack dit...
—Ne parlez pas de Jack. Ce que dit un mari, on le sait d’avance. Qu’est-
ce que vous avez fait?
—J’ai fait une impression durable.
—Et moi aussi... pendant quatre mois. Mais cela ne m’a pas le moins du
monde consolée. J’avais pris l’homme en grippe. Est-ce que vous n’allez
pas cesser de sourire de cet air insondable et me dire où vous voulez en
venir?
Alors mistress Mallowe parla.
. . . . . . . . . . . .
.
—Et vous prétendez soutenir que tout cela fut purement platonique de
part et d’autre?
—Absolument; et dans le cas contraire, je ne m’y serais point
embarquée.
—Et c’est à vous qu’il doit sa dernière promotion?
Mistress Mallowe affirma d’un signe de tête.
—Et vous l’avez mis en garde contre la petite Topsham?
Autre signe affirmatif.
—Et vous lui avez parlé du mémoire particulier envoyé sur son compte
par sir Dugald Delane?
Troisième signe affirmatif.
—Pourquoi?
—Quelle question à faire à une femme? D’abord, parce que cela
m’amusait. Aujourd’hui, je suis fière de ma conquête. Si je vis, il
continuera à réussir. Oui, je le mettrai sur le chemin qui mène tout droit à la
croix de Chevalier, à tout ce qui peut avoir quelque prix aux yeux d’un
homme. Quant au reste, cela le regarde.
—Polly, vous êtes la plus extraordinaire des femmes.
—Pas le moins du monde. Je me concentre, voilà tout. Vous, vous vous
éparpillez, ma chère, et bien que tout Simla connaisse votre habileté à
conduire un attelage...
—Ne pourriez-vous pas employer un terme plus gracieux?
—Un attelage à six, depuis le Mussuck jusqu’au petit Hawley, vous n’y
gagnez rien, pas même de vous amuser.
—Et vous?
—Essayez ma recette. Prenez un homme et non point un gamin, notez
bien, un homme très mûr, sans attaches, et soyez pour lui un guide, un
philosophe, une amie. Vous trouverez là l’occupation la plus intéressante à
laquelle vous vous soyez jamais adonnée. C’est chose possible, vous n’avez
pas besoin de me regarder comme cela—puisque je l’ai fait.
—Il y a là un élément de danger qui donne de l’attrait à l’aventure. Je
chercherai un homme de ce genre et lui dirai: «Maintenant, il est bien
entendu qu’il ne doit pas être question de flirt. Faites exactement ce que je
vous dirai. Mettez à profit mes renseignements et mes conseils, et tout ira
bien.» Est-ce là votre idée!
—Plus ou moins, dit mistress Mallowe avec un sourire énigmatique,
mais assurez-vous que l’on vous comprend.
II
Pan! pan! Pouf! pouf! Quel nuage
de poussière sale. Ma poupée a eu un
accident. Et toute la sciure de bois est
partie.
(Chanson de nourrice.)
Ainsi, à la Fonderie, d’où l’on voit tout le Mail de Simla, mistress
Hauksbee était aux pieds de mistress Mallowe, et recueillait les leçons de la
sagesse.
Le résultat de cet entretien fut la grande idée que mistress Hauksbee était
fière de prendre à son compte.
 —Je vous en avertis, dit mistress Mallowe, qui éprouvait déjà quelques
remords de sa suggestion, la chose n’est pas aussi aisée qu’on le croirait. La
première femme venue,—la petite Topsham elle-même—, conquérir un
homme, mais il en est très peu, très peu qui sachent le manier quand il est
pris.
—Mon enfant, j’ai été un Saint Siméon Stylite féminin, qui regardait
d’en haut les hommes, pendant ces... ces dernières années. Demandez au
Mussuck si je m’entends à les mener.
Mistress Hauksbee s’en alla en chantonnant: j’irai à lui et je lui dirai du
ton le plus ironique, pendant que mistress Mallowe riait toute seule.
Puis, elle devint tout à coup sérieuse:
—Je me demande si j’ai bien fait de conseiller cette distraction. Lucy est
une femme avisée, mais un peu trop étourdie.
Huit jours plus tard, elles se rencontrèrent à un concert du lundi.
—Eh bien? demanda mistress Mallowe.
—Je le tiens, dit mistress Hauksbee, les yeux pétillants de gaieté.
—Qui est-il, petite fille? Je suis désolée de vous avoir parlé de cela.
—Regardez entre les colonnes, au troisième rang, le quatrième à partir
du bout. En ce moment, vous pouvez voir sa figure. Regardez.
—Otis Yeere! C’est bien le dernier auquel j’aurais songé. Je n’aurais
jamais cru! Je ne vous crois pas.
—Ah! Eh bien, attendez que mistress Tarkass ait commencé à démolir
Milton Wellings, alors je vous dirai tout. Chut! la voix de cette femme me
rappelle toujours un train souterrain qui passe sous Earl’s court avec les
freins serrés. Écoutez à présent. Il s’agit vraiment d’Otis Yeere?
—Oui, je le vois, mais il ne s’ensuit pas qu’il soit votre conquête.
—Il l’est, par droit de premier occupant. Je l’ai trouvé abandonné, sans
amis, le soir même de notre entretien, au Burra Khana[C] de Dugald Delane.
Ses yeux m’ont plu et j’ai causé avec lui. Le lendemain il m’a rendu visite.
Le surlendemain, nous avons fait ensemble une promenade à cheval.
Aujourd’hui il est attaché par les mains et les pieds aux roues de ma voiture.
Vous verrez, quand le concert sera fini. Il ne sait pas encore que je suis ici.
—Grâce à Dieu, vous n’avez pas choisi un gamin. Qu’allez-vous faire de
lui, en supposant que vous ayez fait sa conquête!