Endless Ebook, One Click Away – Start Downloading at ebookname.

com

The multivariate social scientist introductory

statistics using generalized linear models
Sofroniou

https://ebookname.com/product/the-multivariate-social-
scientist-introductory-statistics-using-generalized-linear-
models-sofroniou/

OR CLICK HERE

DOWLOAD EBOOK

Browse and Get More Ebook Downloads Instantly at https://ebookname.com

Click here to visit ebookname.com and download ebook now
Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

An Introduction to Generalized Linear Models

Quantitative Applications in the Social Sciences 1st
Edition George Henry Dunteman

https://ebookname.com/product/an-introduction-to-generalized-
linear-models-quantitative-applications-in-the-social-
sciences-1st-edition-george-henry-dunteman/

Generalized Linear Mixed Models Modern Concepts Methods

and Applications 1st Edition Stroup

https://ebookname.com/product/generalized-linear-mixed-models-
modern-concepts-methods-and-applications-1st-edition-stroup-2/

Generalized linear mixed models modern concepts methods

and applications 1st Edition Stroup

https://ebookname.com/product/generalized-linear-mixed-models-
modern-concepts-methods-and-applications-1st-edition-stroup/

Sabine Pass The Confederacy s Thermopylae First Edition

Edward T. Cotham

https://ebookname.com/product/sabine-pass-the-confederacy-s-
thermopylae-first-edition-edward-t-cotham/
Elementary Number Theory 2nd Edition Underwood Dudley

https://ebookname.com/product/elementary-number-theory-2nd-
edition-underwood-dudley/

Annual Reports on NMR Spectroscopy 69 1st Edition

Graham A. Webb (Eds.)

https://ebookname.com/product/annual-reports-on-nmr-
spectroscopy-69-1st-edition-graham-a-webb-eds/

Augustine and the cure of souls revising a classical

ideal 1st Edition Kolbet

https://ebookname.com/product/augustine-and-the-cure-of-souls-
revising-a-classical-ideal-1st-edition-kolbet/

Living History First Printing Edition Hillary Rodham

Clinton

https://ebookname.com/product/living-history-first-printing-
edition-hillary-rodham-clinton/

Metropolitan Art and Literature 1810 1840 Cockney

Adventures Gregory Dart

https://ebookname.com/product/metropolitan-art-and-
literature-1810-1840-cockney-adventures-gregory-dart/
MRI of Tissues with Short T2s or T2 s 1st Edition
Graeme M. Bydder

https://ebookname.com/product/mri-of-tissues-with-
short-t2s-or-t2-s-1st-edition-graeme-m-bydder/
The Multivariate Social Scientist

Introductory Statistics

Using Generalized Linear Models

Copyrighted Material
 To Andrea and Alexandra
Graeme Hutcheson

I;TOV<; 7r. 8E/LUJTOf1,:>-'i)<; f1,m E:>-'u5:vo: X O:T(fJ�wavvov

Nick Sofroniou

Copyrighted Material
The Multivariate Social Scientist

Introductory Statistics

Using Generalized Linear Models

Graeme D. Hutcheson and Nick Sofroniou

SAGE Publications

London ' Thousand Oaks' New Delhi

Copyrighted Material
© Graeme D . Hutcheson and Nick Sofroniou 1999

First published 1999

All rights reserved. No part of this publication may be reproduced, stored in a

retrieval system, transmitted or utilized in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without permission in writing
from the Publishers.

SAG E Publications Ltd

6 Bonhill Street
London EC2A 4PU

SAGE Publication Inc

2455 Teller Road
Thousand Oaks, California 9 1320

SAGE Publications India Pvt Ltd

32, M-Block Market
Greater Kailash - I
New Delhi llO 048

British Library Cataloguing in Publication data

A catalogue record for this book is

available from the British Library

ISBN 0 7619 5200 4

ISBN 0 7619 5201 2 (pbk)

Library of Congress catalog card record available

Typeset by Graeme D. Hutcheson and Nick Sofroniou using �TEX·

Printed in Great Britain by The Cromwell Press Ltd, Trowbridge, Wiltshire

Copyrighted Material
Contents

Preface xi

1 Introduction 1
1 . 1 Generalized Linear Models. 2
1 . 1 . 1 The Random Component . 3
l.l.2 The Systematic Component 3
1 . 1 .3 The Link Function . . . . . 4
l.2 Data Formats . . . . . . . . . . . . 5
l.3 Standard Statistical Analyses within the GLM Framework . 6
l.4 Goodness-of-fit Measures and Model-Building 7
l.5 The Analysis of Deviance . . . . . . . . . . . . . . . . . . . 9
l.6 Assumptions of GLMs . . . . . . . . . . . . . . . . . . . . . 10
l.6. 1 Properties of Scales and Corresponding Transformations 10
l . 6 . 2 Overdispersion and the Variance Function 11
l.6.3 Non-linearity and the Link Function . . . . . . . . . .
. 12
l.6.4 Independent Measurement . . . . . . . . . . . . . . . . . 12
l.6.5 Extrapolating Beyond the Measured Range of Variates . 13
l.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2 Data Screening 15
2 . 1 Levels of Measurement 16
2.2 Data Accuracy . . . 17
2.3 Reliable Correlations 17
2 . 4 Missing Data . . . . 18
2 .5 Outliers . . . . . . . 19
2 .5. 1 On a Single Variable 20
2.5.2 Across Multiple Variables 20
2.5.3 Identifying Outliers 21
Leverage Values 21
Cook's Distance . . 22
2.5.4 Dealing with Outliers 24
2.6 Using Residuals to Check for Violations of Assumptions 25
2 . 6 . 1 Ordinary Residuals . . . . . . . . . . . . . . . . . 26

Copyrighted Material
vi CONTENTS

2.6.2 Standardized Residuals 26

2.6.3 Studentized Residuals 26
2.7 Assumptions . . . . . . . 26
2 .7.1 Normality . . . . . 26
2.7.2 Constant Variance 28
2.7.3 Linearity . . . . . 29
2.7.4 Independence . . . 31
2.8 Transformation and Curvilinear Models 31
2 . 8 . 1 Box-Cox Transformation with Skewed Errors 32
2.8.2 Fitting Models to Curvilinear Data Trends . 35
2.9 Multicollinearity and Singularity . . . . . . . . . . . 41
2 . 10 Diagnostics for Logistic Regression and Loglinear Models 42
2 . 1 0 . 1 Under and Overdispersion . . . . . . . . . . . . . . 42
2.11 Summary and Recommendations on the Order of Screening
Practices . . . . . . . . . . . . . 44
2.12 Statistical Software Commands 45
2 . 1 2 . 1 SPSS 45
2. 1 2 . 2 GLIM . . . . . . . . .
. 49

3 Ordinary Least-Squares Regression 55

3 . 1 Simple OLS Regression . . . . . 56
3. 1 . 1 The Regression Equation . 56
3. 1 .2 Confidence Intervals for {3 . 59
3. 1 .3 Confidence Intervals for Fitted y 62
3 . 1 . 4 Prediction Intervals . . . 63
3. 1.5 Goodness-of-fit Measures 65
The R2 Statistic . . . . . 65
The F Statistic . . . . . . 66
3.2 A Worked Example of Simple Regression . 68
3.3 Multiple OLS Regression . . . . . . . . 71
3.3 . 1 The Regression Equation . . . 73
Interactions and Curvilinearity 73
3.3.2 Confidence Intervals for {3 . . . 75
3.3.3 Confidence Intervals for Fitted y 75
3.3.4 Prediction Intervals . , . 75
3.3.5 Goodness-of-fit Measures . . . 76
The R2 and R; Statistics . . . 76
The F and Partial-F Statistics 77
3.3.6 Multicollinearity . . . . 78
Perfect Multicollinearity . . . . 79
High Multicollinearity . . . . . 80
Identifying Instances of Multicollinearity . 82
Dealing with Multicollinearity . . . . . . . 84
3.3.7 Dummy Variable Coding Discontinuous Data 85
Indicator Coding . . . . . . . . . . . . . . . . 86

Copyrighted Material
CONTENTS vii

Deviation Coding . . . . . . . . . . . . . . . 90
Dummy Coding Ordered Categorical Data . 93
3.3.8 Model Selection . . . . . . . . . . . . . . . . 94
Criteria for Including and Removing Variables 94
3.3.9 Automated Model Selection 95
Forward Selection . . 96
Backward Elimination . . . 97
Stepwise Selection . . . . . 97
3.4 A Worked Example of Multiple Regression . 98
3.5 Summary . . . . . . . . . . . . 101
3.6 Statistical Software Commands 102
3.6.1 SPSS 102
3.6.2 GLIM . . 106

4 Logistic Regression 113

4 . 1 Simple Logistic Regression . . . . . . . . . . . . . . . . . 114
4 . 1 . 1 The Regression Equation . . . . . . . . . . . . . 1 17
Calculating and Interpreting Model Parameters . 121
4 . 1 .2 Confidence Intervals for f3 . . . . 1 23
4. 1 .3 Confidence Intervals for Fitted y 1 24
4 . 1 . 4 Under and Overdispersion . . 1 27
4 . 1 .5 Goodness-of-fit Measures . . 128
The Log-Likelihood Statistic 128
The Wald Statistic . . . . . . 129
Analysis of Classification Table Data . 1 29
4.2 A Worked Example of Simple Logistic Regression . 1 30
4.3 Multiple Logistic Regression . . . . . . 134
4.3.1 The Regression Equation . . . . . . . . .
. 134
Interactions and Curvilinearity . . . . . . . 1 35
Calculating and Interpreting Model Parameters . 136
4.3.2 Confidence Intervals for f3 . . . . 1 37
4.3.3 Confidence Intervals for Fitted y 1 38
4.3.4 Goodness-of-fit Measures . . 1 39
The Log-Likelihood Statistic . . 139
4.3.5 Multicollinearity . . . . . . . . . 141
4.3.6 Dummy Variable Coding Discontinuous Data 141
Indicator Coding . 143
Deviation Coding . . . . . . . . . . . . . . . . 144
4.3.7 Model Selection . . . . . . . . . . . . . . . . . 1 45
Criteria for Including and Removing Variables 1 46
4.4 A Worked Example of Multiple Logistic Regression . 1 47
4.5 Summary . . . . . . . . . . . . 152
4.6 Statistical Software Commands 152
4 . 6 . 1 SPSS 152
4.6.2 GLIM . . . . . . . . . . 156

Copyrighted Material
V III CONTENTS

5 Loglinear Analysis 163

5.1 Traditional Bivariate Methods . . . . . . . . . . . . 164
5.1 . 1 Hypothesis Testing with Significance Tests 1 64
5.1 .2 Goodness-of-fit Statistics 1 65
The Pearson X 2 Statistic 1 65
The C2 Statistic . . . . . 1 66
Adjusted Residuals . . . . 1 67
5.1.3 Modelling with Sample Odds Ratios 1 68
Confidence Intervals . . 1 69
5.2 Loglinear Models . . . . . . . . 170
5.2.1 The Loglinear Equation 171
5.2.2 A Bivariate Example . . 172
5.2.3 A Multivariate Example 175
5.2.4 Model Notation . . . . . 175
5.2.5 Model Building . . . . . 177
5.2.6 Goodness-of-fit Statistics and Residuals 179
5.2.7 Explanatory Variables with Ordered Categories . 1 82
Scoring Ordered Categorical Variables . . . . . . 1 83
Bivariate Example with Explanatory Variables having
Ordered Categories . . . . . . . . . . . . . . . . . . . . 184
5.2.8 Associations Between Nominal and Ordered
Categorical Variables . . . . . . . . . . . . . . . . . . . . 186
5.2.9 Homogeneous Linear x Linear Associations . . . . . . . 189
.

5.2.10 Logit Analysis - Evaluating a Binary Response Variable 1 9 1

5.2.11 Multicategory Response Variables with Nominal
and Ordered Categories . . . . . . . . . . . . . . . . . . 1 92
Nominal Response Variables: Baseline Category Logits . 192
Ordered Response Variables . 1 93
5.3 Summary . . . . . . . . . . . . 199
5.4 Statistical Software Commands 200
5.4.1 SPSS 200
5.4.2 GLIM 205

6 Factor Analysis 217

6 . 1 Overview . . . . . . . . . 218
6.2 Factor Analysis Equations 220
6.3 Preliminary Analyses . . . 221
6.3.1 Data Assumptions 222
Continuous and Ordered Categorical Data . 222
Sample Size . . . . . . . 222
6.3.2 Data Screening . . . . . . . . . . 222
6.3.3 Examining Correlations . . . . . 223
6.3.4 Measures of Sampling Adequacy 224
6.4 Factor Extraction . . . . . . . . . . . . 226
6.4.1 Principal Components Analysis . 226

Copyrighted Material
CONTENTS IX

Selecting the Number of Factors 228

6.4.2 Interpreting the Initial Factors . 230
6.5 Factor Rotation . . . . . . . . . . . . . . 232
6.5.1 Interpreting the Rotated Factors 235
6.6 Including Factors in GLMs . . . . . . 237
6.7 A Worked Example of Factor Analysis 238
6.8 Summary . . . . . . . . . . . . 248
6.9 Statistical Software Commands 248
6.9.1 SPSS . . . . . . . . . . 249

7 Conclusion 253
7. 1 Main Points of the GLM Framework 253
7.2 Sampling Assumptions and GLMs . . . . . . . . . . . . . . . . 254
7.3 Measurement Assumptions of Explanatory Variables in GLMs . 255
7.4 Ordinal Variables . . . . . . . . . . . . . . 256
7.5 GLM Variants of ANOVA and ANCOVA 257
7.6 Repeated Measurements . . . . . . . . 258
7.7 Time Series Analysis . . . . . . . . . . 259
7.8 Gamma Errors and the Link Function: an Alternative to
Data Transformation . . . . . . . . . . 260
7.9 Survival Analysis . . . . . . . . . . . . 261
7.10 Exact Sample and Sparse Data Methods 262
7. 1 1 Cross Validation of Models 263
7 . 1 2 Software Recommendations 263
7.12. 1 SPSS for Windows 263
7. 12.2 GLIM . . . . . . . . 264
7.12.3 SAS . . . . . . . . . 264
7.12.4 GENSTAT for Windows 264
7.12.5 Overall . . . . . . . . . 265

References 267

Index 274

Copyrighted Material
Copyrighted Material
Preface

One of the most important contributions to the field of statistics in the latter
part of this century has been the introduction of the concept of generalized lin­
ear models by J. Neider and R. W. M. Wedderburn in 1 972. This framework
unifies the modelling techniques for categorical data, such as logistic regression
and loglinear models, with the traditional linear regression and ANOVA meth­
ods. Within one paradigm we have both an integrated conceptual framework
and an emphasis on the explicit analysis of data through a model-building
approach allowing the estimation of the size of effects, predictions of the re­
sponse variable, and the construction of confidence intervals. This move away
from the simplistic hypothesis testing framework that has come to characterize
much of social science research, with its binary conception of 'scientific truth',
can only be for the better - allowing the researcher to focus on the practical
importance of a given variable in their particular domain of interest.
It is an unfortunate fact that the widespread assimilation of these methods
into the social sciences is long overdue, and this was one of the motivations
behind the decision of the authors to undertake the sometimes arduous task
of writing this book. An additional reason was the lack of texts that make use
of the common theoretical underpinnings of generalized linear models ( GLMs )
as a teaching tool. So often one finds that introductory textbooks continue
to present statistical techniques as disparate methods with little cohesion. In
contrast we have begun with a brief exposition of the common conceptual
framework and written the subsequent descriptions of the methods around
this, making explicit the relationships between them. Our experience has
been that students benefit from this unity and the insight which follows the
extension of the model specification, criticism and interpretation techniques,
learned with continuous data, to binary and multi-category data.
In keeping with the attempt to integrate modern statistics into social sci­
ence research we have chosen to replace certain archaic terminology with cur­
rent statistical equivalents. Thus, the terms explanatory and response variable
are used instead of independent and dependent variable. Similarly, we have
encouraged the use of the terms categorical variable, unordered or ordered,
and continuous variable which more neatly map onto the GLM framework
than the traditional classifications of variables into nominal, ordinal, interval

xi

Copyrighted Material
XII Preface

and ratio scales. The terminology used in this book has been standardized
to follow that used in McCullagh and Neider ( 1 989) and Agresti ( 1990 ) , two
definitive books on the methods described here.
The data sets used in the present book were developed during the teach­
ing of these methods over a number of years, most are hypothetical and are
designed to illustrate the range of techniques covered. It is hoped that the
use of made-up sets will encourage readers to experiment with the data - to
change distributions and add variables, examining the effects upon the model
fit. References are provided for examples with real-life data sets.
We have tried to make the chapters as software-independent as possible, so
that the book can be used with a wide variety of statistical software packages.
For this reason, the commands for two packages we have used for illustration
are confined to an appendix at the end of each chapter. SPSS for Windows
was chosen because of its friendly interface and widespread availability, whilst
GLIM was selected because of its sophisticated model specification syntax and
its integrated presentation of generalized linear modelling methods.
The material covered in the book has been taught as a course at Glasgow
University where it was presented to post-graduate students and researchers
( 1 994�1 998) . Each chapter can be taught in two 2-hour sessions, making a
complete course of ten 2-hour sessions viable. All data sets and GLIM code
are reproduced in the book and can also be obtained from the StatLib site on
the World Wide Web at: http://lib/stat/cmu.edu/datasets/.
We would like to express our thanks to the anonymous reviewers for Sage
P ublications who gave many helpful comments on our earlier drafts. We would
also like to express our gratitude to those who have provided a number of
excellent courses in categorical methods and GLMs, particularly Richard B .
Davies, Damon M. Berridge, and Mick Green o f Lancaster University, and
Ian Diamond of Southampton University. Murray Aitkin kindly sent pre­
publication copies of his article and macros for hierarchical modelling in GLIM,
and James K . Lindsey obtained a copy for us of his unfortunately out of print
text on the Analysis of Stochastic Processes using GLIM ( 1 992) .
This book was typeset by the authors using g\1EX- and a debt of gratitude is
owed to Donald Knuth, the creator of 1EX- (Knuth, 1984) , Leslie Lamport who
built this into the g\1EX- document preparation system (Lamport, 1 994) , and
to the many contributors who freely give their time and expertise to support
this package.

Copyrighted Material
Chapter 1

Introduction

For many years the social sciences have been characterized by a restricted
approach to the application of statistical methods. These have emphasized
the use of analyses based on the assumptions of the parametric family of
statistics, particularly for multivariate data. Where methods have been used
making less restrictive assumptions, these have typically been limited to data
with only two variables, e.g . , the non-parametric techniques described by Siegel
and Castellan ( 1988) . There has also been an emphasis on hypothesis testing
using the convention of statistical significance at the P ::; 0.05 level as a
criterion for the inclusion of a variable in one's theoretical framework. This
approach to 'scientific truth' is detrimental in that it restricts the analysis
which can be applied to the data, and does not distinguish between statistical
and substantive significance. The distinction between the two is important
as significance in one does not necessarily indicate significance in the other.
Statistical significance indicates whether a particular result is likely to have
arisen by chance ( the criterion of 0.05 being a convenient convention ) , whereas
substantive significance indicates the practical importance of a given variable
or set of variables in the field of interest.
In contrast to this approach to data analysis a suite of statistical tech­
niques have been developed, which have been applied mainly in the biological
and medical fields, and offer multiple-variable techniques for dealing with data
that do not meet all the requirements of traditional parametric statistics, such
as binary and frequency data. Alongside this wider development of statistics
has been an emphasis on model building rather than on mere hypothesis testing
with greater use of confidence intervals to enable the predictive utility of mod­
els to be estimated in addition to their statistical significance. One important
consequence of these developments has been the unification of traditional para­
metric techniques with methods for data which depart from linear parametric
data assumptions through the conceptual framework provided by Generalized
Linear Models ( GLMs ) . This unified view has allowed us to organize this book
around variations of a single theme - that by examining the properties of a

Copyrighted Material
2 Introduction

data set, one can choose from a range of GLMs to develop a model of the data
that offers both parsimony of explanation and a measure of the model's utility
for prediction purposes. Thus, two common goals of science, the explanation
and prediction of phenomena, may be successfully developed within the social
sciences.
Whilst some of the statistical methods we shall describe have been devel­
oped within the framework of quantitative sociology, e.g., Goodman ( 1970;
1 979), Bishop, Fienberg and Holland ( 1 975) , and Clogg ( 1 982) , it is our opin­
ion that the widespread application of this approach to data from the social
sciences is long overdue. In presenting this statistical framework, it has been
assumed that the reader has completed an undergraduate course in statis­
tics and is familiar with the concepts of linear regression, analysis of variance
(ANOVA), and the analysis of contingency tables using the Pearson X2 statis­
tic.

1.1 Generalized Linear Mo dels

The term Generalized Linear Model, refers to a family of statistical models that
extend the linear parametric methods such as ordinary least-squares (OLS)
regression and analysis of variance, to data types where the response variable
is discrete, skewed, and/or non-linearly related to the explanatory variables.
GLMs seek to explain variation in a response variable of interest in terms of
one or more explanatory variables. The GLM modelling scheme was originally
developed by Neider and Wedderburn ( 1972) and extended in McCullagh and
NeIder ( 1 989) and can be summarized as having three components, a random
component, a systematic component and a function which links the two.

1. The random component is the probability distribution assumed to

underlie the response variable. The basic assumption is that a given data
set is a random sample of independent observations, though variations
exist to take into account observations that are not independent.

2. The systematic component is the fixed structure of the explanatory

variables, usually characterized by a linear function.

3. The link function maps the systematic component onto the random
component. This function can be one of identity for Normally distributed
random components, or one of a number of non-linear links when the
random component is not Normally distributed.

The terms 'multivariate' and 'multiple' regression are sometimes used in­
terchangeably, which is unfortunate, since they have different meanings. 'Mul­
tiple' regression refers to an analysis with a single response variable and sev­
eral explanatory variables, and is a univariate statistical technique. Whilst, a
'multivariate' statistical test refers to an analysis with more than one response

Copyrighted Material
 Another Random Document on
Scribd Without Any Related Topics
The Project Gutenberg eBook of The Old Red
Sandstone; or, New Walks in an Old Field
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: The Old Red Sandstone; or, New Walks in an Old Field

Author: Hugh Miller

Release date: November 30, 2020 [eBook #63923]

Most recently updated: October 18, 2024

Language: English

Credits: Produced by Tom Cosmas from files kindly provided by The

Internet Archive

*** START OF THE PROJECT GUTENBERG EBOOK THE OLD RED

SANDSTONE; OR, NEW WALKS IN AN OLD FIELD ***
 Note: Click on all images to view larger version.

THE

OLD RED SANDSTONE;

OR,

NEW WALKS IN AN OLD FIELD.

BY
 HUGH MILLER,
AUTHOR OF "FOOT-PRINTS OF THE CREATOR," ETC.
ILLUSTRATED WITH NUMEROUS ENGRAVINGS.
FROM THE FOURTH LONDON EDITION.
BOSTON:
GOULD AND LINCOLN,
59 WASHINGTON STREET.
1851.
STEREOTYPED AT THE BOSTON STEREOTYPE FOUNDRY.
Printed by G. C. Rand & Co. No. 3 Cornhill.
 DEDICATION.
TO

RODERICK IMPEY MURCHISON,

Esq., F. R. S., Etc.,
PRESIDENT OF THE GEOLOGICAL SOCIETY.
In the autumn of last year, I sat down to write a few geological
sketches for a newspaper; the accumulated facts of twenty years
crowded upon me as I wrote, and the few sketches have expanded
into a volume. Permit me, honored Sir, to dedicate this volume to
you. Its imperfections are doubtless many, for it has been produced
under many disadvantages; but it is not the men best qualified to
decide regarding it whose criticisms I fear most; and I am especially
desirous to bring it under your notice, as of all geologists the most
thoroughly acquainted with those ancient formations which it
professes partially to describe. I am, besides, desirous it should be
known, and this, I trust, from other motives than those of vanity,
that, when prosecuting my humble researches in obscurity and
solitude, the present President of the Geological Society did not
deem it beneath him to evince an interest in the results to which
they led, and to encourage and assist the inquirer with his advice.
Accept, honored Sir, my sincere thanks for your kindness.
Smith, the father of English Geology, loved to remark that he had
been born upon the Oolite—the formation whose various deposits he
was the first to distinguish and describe, and from which, as from
the meridian line of the geographer, the geological scale has been
graduated on both sides. I have thought of the circumstance when,
on visiting in my native district the birthplace of the author of the
Silurian System, I found it situated among the more ancient
fossiliferous rocks of the north of Scotland—the Lower Formation of
the Old Red Sandstone spreading out beneath and around it, and
the first-formed deposit of the system, the Great Conglomerate,
rising high on the neighboring hills. It is unquestionably no slight
advantage to be placed, at that early stage of life, when the mind
collects its facts with greatest avidity, and the curiosity is most
active, in localities where there is much to attract observation that
has escaped the notice of others. Like the gentleman whom I have
now the honor of addressing, I too was born on the Old Red
Sandstone, and first broke ground as an inquirer into geological fact
in a formation scarce at all known to the geologist, and in which
there still remains much for future discoverers to examine and
describe. Hence an acquaintance, I am afraid all too slight, with
phenomena which, if intrinsically of interest, may be found to have
also the interest of novelty to recommend them, and with organisms
which, though among the most ancient of things in their relation to
the world's history, will be pronounced new by the geological reader
in their relation to human knowledge. Hence, too, my present
opportunity of subscribing myself, as the writer of a volume on the
Old Red Sandstone,
Honored Sir,

With sincere gratitude and respect,

Your obedient humble Servant,

HUGH MILLER.
Edinburgh, May 1, 1841.
 PREFACE.
Nearly one third of the present volume appeared a few months
ago in the form of a series of sketches in the Witness newspaper. A
portion of the first chapter was submitted to the public a year or two
earlier, in Chambers's Edinburgh Journal. The rest, amounting to
about two thirds of the whole, appears for the first time.
Every such work has its defects. The faults of the present volume
—faults all too obvious, I am afraid—would have been probably
fewer had the writer enjoyed greater leisure. Some of them,
however, seem scarce separable from the nature of the subject:
there are others for which, from their opposite character, I shall have
to apologize in turn to opposite classes of readers. My facts would,
in most instances, have lain closer had I written for geologists
exclusively, and there would have been less reference to familiar
phenomena. And had I written for only general readers, my
descriptions of hitherto undescribed organisms, and the deposits of
little-known localities, would have occupied fewer pages, and would
have been thrown off with, perhaps, less regard to minute detail
than to pictorial effect. May I crave, while addressing myself, now to
the one class, and now to the other, the alternate forbearance of
each?
Such is the state of progression in geological science, that the
geologist who stands still for but a very little, must be content to
find himself left behind. Nay, so rapid is the progress, that scarce a
geological work passes through the press in which some of the
statements of the earlier pages have not to be modified, restricted,
or extended in the concluding ones. The present volume shares, in
this respect, in what seems the common lot. In describing the
Coccosteus, the reader will find it stated that the creature, unlike its
contemporary the Pterichthys, was unfurnished with arms. Ere
arriving at such a conclusion, I had carefully examined at least a
hundred different Coccostei; but the positive evidence of one
specimen outweighs the negative evidence of a hundred; and I have
just learned from a friend in the north, (Mr. Patrick Duff, of Elgin,)
that a Coccosteus lately found at Lethen-bar, and now in the
possession of Lady Gordon Gumming, of Altyre, is furnished with
what seem uncouth, paddle-shaped arms, that project from the
head.[A] All that I have given of the creature, however, will be found
true to the actual type; and that parts should have been omitted will
surprise no one who remembers that many hundred belemnites had
been figured and described ere a specimen turned up in which the
horny prolongation, with its enclosed ink-bag, was found attached to
the calcareous spindle; and that even yet, after many thousand
trilobites have been carefully examined, it remains a question with
the oryctologist, whether this crustacean of the earliest periods was
furnished with legs, or creeped on an abdominal foot, like the snail.
[A] As these paddle-shaped arms have not been introduced by Agassiz into his
restoration of the Coccosteus, their existence, at least as arms, must still be
regarded as problematical. There can be no doubt, however, that they existed as
plates of very peculiar form, and greatly resembling paddles, and that they served
in the economy of the animal some still unaccounted for purpose.

I owe to the kindness of Mr. Robertson, Inverugie, the specimen

figured in Plate V., fig. 7, containing shells of the only species yet
discovered in the Old Red Sandstone of Scotland. They occur in the
Lower Formation of the system, in a quarry near Kirkwald, in which
the specimen figured, with several others of the same kind, was
found by Mr. Robertson, in the year 1834. In referring to this shell,
page 99,[B] I have spoken of it as a delicate bivalve, much
resembling a Venus; drawing my illustration, naturally enough, when
describing the shell of an ocean deposit, rather from among marine,
than fluviatile testacea. I have since submitted it to Mr. Murchison,
who has obligingly written me that he "can find no one to say more
regarding it than that it is very like a Cyclas." He adds, however, that
it must be an ocean production notwithstanding, seeing that all its
contemporaries in England, Scotland, and Russia, whether shells or
fish, are unequivocally marine.
[B] Page 90 of the present edition.

With the exception of two of the figures in Plate IX., the figures
of the Cephalaspis and the Holoptychius, and one of the sections in
the Frontispiece, section 2, all the prints of the volume are originals.
To Mr. Daniel Alexander, of Edinburgh,—a gentleman, who to the skill
and taste of the superior artist, adds no small portion of the
knowledge of the practical geologist,—I am indebted for several of
the drawings; that of fig. 2 in Plate V., fig. 1 in Plate VI., fig. 2 in
Plate VIII., and figs. 3 and 4 in plate IX. I am indebted to another
friend for fig. 1, in Plate VII. Whatever defects may be discovered in
any of the others, must be attributed to the untaught efforts of the
writer, all unfamiliar, hitherto, with the pencil, and with by much too
little leisure to acquaint himself with it now.
 AMERICAN PUBLISHERS' NOTICE.
The publishers take pleasure in presenting to the American
reader this interesting work of Hugh Miller, in which are restored to
our view some of the phenomena which occurred in the earlier
formations of the crust of the earth, belonging to those
inconceivably remote ages when living things first appeared;—a
work so scientific, and yet so illustrated with familiar objects and
scenes, as to be well understood by those little versed in Geology.
The grand conclusions which the author deduces from apparently
trifling circumstances that every one has noticed a hundred times,
without being the wiser, illustrate the difference between the
philosopher and the common observer; and the simple and pictorial
style in which they are delineated renders the work peculiarly
fascinating.
This is a reprint of the fourth English edition, without additions or
alterations, excepting the omission of the prefatory Notes to the
second and third editions. In the first of these, the author states that
he had added about fifteen pages to the first edition, chiefly relating
to that middle formation of the system to which the organisms of
Balruddery and Carmylie belong, the representative of the
Cornstones in England. Some matters there given as merely
conjectural were also replaced by ascertained facts. In the latter, he
announces that the somewhat bold prediction made by him in the
first edition, in 1841, that the ichthyolites of the Old Red Sandstone
would be found at least equal to those of all the geological
formations united, at the death of Cuvier, was already more than
fulfilled. Cuvier enumerated ninety-two species of fossil fishes;
Agassiz, in 1846, enumerated one hundred and five in the Old Red
Sandstone alone, a formation which had been regarded as poorer in
organisms than any other. In this edition was given the list of
species, as determined and arranged by Professor Agassiz. Many
additions in the shape of notes were also made.
In the first two editions it was stated that there was a gradual
increase of size observable in the progress of ichthyolic life, and that
the Old Red System exhibited, in its successive formations, this
gradation of bulk, beginning with an age of dwarfs, and ending with
an age of giants. Since then, it has been ascertained that there were
giants among the dwarfs. The remains of one of the largest fish
found any where, has been discovered in its lowest formation;
whereby he was convinced that the theory of a gradual progression
in size, from the earlier to the later Palæozoic formations, though
based originally on no inconsiderable amount of negative evidence,
must be permitted to drop. On this fact he has based his
incontrovertible argument against the "development theory" in his
more recent work, already given to the American Public, "Foot-Prints
of the Creator."
Boston, January, 1851.
 CONTENTS.
CHAPTER I.
PAGE
The Working-man's true Policy.—His only Mode of acquiring Power.—
The Exercise of the Faculties essential to Enjoyment.—No necessary
Connection between Labor and Unhappiness.—Narrative.—Scenes in a
Quarry.—The two dead Birds.—Landscape.—Ripple Markings on a
Sandstone Slab.—Boulder Stones.—Inferences derived from their water-
worn Appearance.—Sea-coast Section.—My first discovered Fossil.—Lias
Deposit on the Shores of the Moray Frith.—Belemnite.—Result of the
Experience of half a Lifetime of Toil.—Advantages of a Wandering
Profession in Connection with the Geology of a Country.—Geological
Opportunities of the Stone-Mason.—Design of the present Work,

1-14

CHAPTER II.
The Old Red Sandstone.—Till very lately its Existence as a distinct
Formation disputed.—Still little known.—Its great Importance in the
Geological Scale.—Illustration.—The North of Scotland girdled by an
immense Belt of Old Red Sandstone.—Line of the Girdle along the Coast.
—Marks of vast Denudation.—Its Extent partially indicated by Hills on the
western Coast of Ross-shire.—The System of great Depth in the North of
Scotland.—Difficulties in the Way of estimating the Thickness of Deposits.
—Peculiar Formation of Hill.—Illustrated by Ben Nevis.—Caution to the
Geological Critic.—Lower Old Red Sandstone immensely developed in
Caithness.—Sketch of the Geology of that County.—Its strange Group of
Fossils.—their present Place of Sepulture.—Their ancient Habitat.—
Agassiz.—Amazing Progress of Fossil Ichthyology during the last few
Years.—Its Nomenclature.—Learned Names repel unlearned Readers.—
Not a great deal in them,

15-34
 CHAPTER III.
Lamarck's Theory of Progression illustrated.—Class of Facts which give
Color to it.—The Credulity of Unbelief.—M. Maillet and his Fish-birds.—
Gradation not Progress.—Geological Argument.—The Present incomplete
without the Past.—Intermediate Links of Creation.—Organisms of the
Lower Old Red Sandstone.—The Pterichthys.—Its first Discovery.—Mr.
Murchison's Decision regarding it.—Confirmed by that of Agassiz.—
Description.—The several Varieties of the Fossil yet discovered.—Evidence
of violent Death in the Attitudes in which they are found.—The
Coccosteus of the Lower Old Red.—Description.—Gradations from
Crustacea to Fishes.—Habits of the Coccosteus.—Scarcely any Conception
too extravagant for Nature to realize,

35-54

CHAPTER IV.
The Elfin-fish of Gawin Douglas.—The Fish of the Old Red Sandstone
scarcely less curious.—Place which they occupied indicated in the present
Creation by a mere Gap.—Fish divided into two great Series, the Osseous
and Cartilaginous.—Their distinctive Peculiarities.—Geological Illustration
of Dr. Johnson's shrewd Objection to the Theory of Soame Jenyns.—
Proofs of the intermediate Character of the Ichthyolites of the Old lied
Sandstone.—Appearances which first led the Writer to deem it
intermediate.—Confirmation by Agassiz.—The Osteolepis.—Order to
which, this Ichthyolite belonged.—Description.—Dipterus.—Diplopierus.—
Cheirolepis.—Glyptolepis,

55-78

CHAPTER V.
The Classifying Principle and its Uses.—Three Groups of Ichthyolites
among the Organisms of the Lower Old Red Sandstone.—Peculiarities of
the Third Group.—Its Varieties.—Description of the Cheiracanthus.—Of
two unnamed Fossils of the same Order.—Microscopic Beauty of these
ancient Fish.—Various Styles of Ornament which obtain among them.—
The Molluscs of the Formation.—Remarkable chiefly for the Union of
modern with ancient Forms which they exhibit.—Its Vegetables.—
Importance and Interest of the Record which it furnishes,
 79-94

CHAPTER VI.
The Lines of the Geographer rarely right Lines.—These last, however,
always worth looking at when they occur.—Striking Instance in the Line of
the Great Caledonian Valley.—Indicative of the Direction in which the
Volcanic Agencies have operated.—Sections of the Old Red Sandstone
furnished by the granitic Eminences of the Line.—Illustration.—Lias of the
Moray Frith.—Surmisings regarding its original Extent.—These lead to an
exploratory Ramble.—Narrative.—Phenomena exhibited in the Course of
half an Hour's Walk.—The little Bay.—Its Strata and their Organisms,

95-108

CHAPTER VII.
Further Discoveries of the Ichthyolite Beds.—Found in one Locality
under a Bed of Peat.—Discovered in another beneath an ancient Burying-
ground.—In a third underlying the Lias Formation.—In a fourth
overtopped by a still older Sandstone Deposit.—Difficulties in ascertaining
the true Place of a newly-discovered Formation.—Caution against drawing
too hasty Inferences from the mere Circumstance of Neighborhood.—The
Writer receives his first Assistance from without.—Geological Appendix of
the Messrs. Anderson, of Inverness.—Further Assistance from the
Researches of Agassiz.—Suggestion.—Dr. John Malcolmson.—His
extensive Discoveries in Moray.—He submits to Agassiz a Drawing of the
Pterichthys.—Place of the Ichthyolites in the Scale at length determined.
—Two distinct Platforms of Being in the Formation to which they belong,

109-124

CHAPTER VIII.
Upper Formations of the Old Red Sandstone.—Room enough, for each
and to spare.—Middle, or Cornstone Formation.—The Cephalaspis its
most characteristic Organism.—Description.—The Den of Balruddery
richer in the Fossils of this middle Formation than any other Locality yet
discovered.—Various Contemporaries of the Cephalaspis.—Vegetable
Impressions.—Gigantic Crustacean.—Seraphim.—Ichthyodorulites.—
Sketch of the Geology of Forfarshire.—Its older Deposits of the Cornstone
Formation.—The Quarries of Carmylie.—Their Vegetable and Animal
Remains.—The Upper Formation.—Wide Extent of the Fauna and Flora of
the earlier Formations.—Probable Cause,

125-150

CHAPTER IX.
Fossils of the Upper Old Red Sandstone much more imperfectly
preserved than those of the Lower.—The Causes obvious.—Difference
between the two Groups, which first strikes the Observer, a difference in
size.—The Holoptychius a characteristic Ichthyolite of the Formation.—
Description of its huge Scales.—Of its Occipital Bones, Fins, Teeth, and
general Appearance.—Contemporaries of the Holoptychius.—Sponge-like
Bodies.—Plates resembling those of the Sturgeon.—Teeth of various
forms, but all evidently the teeth of fishes.—Limestone Band and its
probable Origin.—Fossils of the Yellow Sandstone.—the Pterichthys of
Dura Den.—Member of a Family peculiarly characteristic of the System.—
No intervening Formation between the Old Red Sandstone and the Coal
Measures.—The Holoptychius contemporary for a time with the
Megalichthys,—The Columns of Tubal-Cain,

151-172

CHAPTER X.
Speculations in the Old Red Sandstone, and their Character.—George,
first Earl of Cromarty.—His Sagacity as a Naturalist at fault in one
instance.—Sets himself to dig for Coal in the Lower Old Red Sandstone.—
Discovers a fine Artesian Well.—Value of Geological Knowledge in an
economic view.—Scarce a Secondary Formation in the Kingdom in which
Coal has not been sought for.—Mineral Springs of the Old Red Sandstone.
—Strathpeffer.—Its Peculiarities whence derived.—Chalybeate Springs of
Easter Ross and the Black Isle.—Petrifying Springs.—Building-Stone and
Lime of the Old Red Sandstone.—Its various Soils,

173-189

CHAPTER XI.
Geological Physiognomy.—Scenery of the Primary Formations; Gneiss,
Mica Schist, Quartz Rock.—Of the Secondary; the Chalk Formations, the
Oolite, the New Red Sandstone, the Coal Measures.—Scenery in the
Neighborhood of Edinburgh.—Aspect of the Trap Rocks.—The Disturbing
and Denuding Agencies.—Distinctive Features of the Old Red Sandstone.
—Of the Great Conglomerate.—Of the Ichthyolite Beds.—The Burn of
Eathie.—The Upper Old Red Sandstones.—Scene in Moray,

190-210

CHAPTER XII.
The two Aspects in which Matter can be viewed; Space and Time.—
Geological History of the Earlier Periods.—The Cambrian System.—Its
Annelids.—The Silurian System.—Its Corals, Encrinites, Molluscs, and
Trilobites.—Its Fish.—These of a high Order, and called into Existence
apparently by Myriads.—Opening Scene in the History of the Old Red
Sandstone a Scene of Tempest.—Represented by the Great
Conglomerate.—Red a prevailing Color among the Ancient Rocks
contained in this Deposit.—Amazing Abundance of Animal Life.—
Exemplified by a Scene in the Herring Fishery.—Platform of Death.—
Probable Cause of the Catastrophe which rendered it such,

211-225

CHAPTER XIII.
Successors of the exterminated Tribes.—The Gap slowly filled.—Proof
that the Vegetation of a Formation may long survive its Animal Tribes.—
Probable Cause.—Immensely extended Period during which Fishes were
the Master-existences of our Planet.—Extreme Folly of an Infidel
Objection illustrated by the Fact.—Singular Analogy between the History
of Fishes as Individuals and as a Class.—Chemistry of the Lower
Formation.—Principles on which the Fish-enclosing Nodules were
probably formed.—Chemical Effect of Animal Matter in discharging the
Color from Red Sandstone.—Origin of the prevailing tint to which the
System owes its Name.—Successive Modes in which a Metal may exist.—
The Restorations of the Geologist void of Color.—Very different
Appearance of the Ichthyolites of Cromarty and Moray,

226-242

CHAPTER XIV.
 The Cornstone Formation and its Organisms.—Dwarf Vegetation.—
Cephalaspides.—Huge Lobster.—Habitats of the existing Crustacea.—No
unapt representation of the Deposit of Balruddery, furnished by a land-
locked Bay in the neighborhood of Cromarty.—Vast Space occupied by the
Geological Formations.—Contrasted with the half-formed Deposits which
represent the existing Creation.—Inference.—The formation of the
Holoptychius.—Probable origin of its Siliceous Limestone.—Marked
increase in the Bulk of the Existences of the System.—Conjectural Cause.
—The Coal Measures.—The Limestone of Burdie House Conclusion,

243-259
Ichthyolites of the Old Red Sandstone—from Agassiz's "Poissons
Fossiles,"

261-288

EXPLANATIONS OF THE SECTIONS AND

PLATES.
SECTION I.
Represents the Old Red System of Scotland from its upper beds of
Yellow Quartzose Sandstone to its Great Conglomerate base. a.
Quartzose Yellow Sandstone, b. Impure concretionary limestone
enclosing masses of chert, c. Red and variegated sandstones and
conglomerate. These three deposits constitute an upper formation of
the system, characterized by its peculiar group of fossils. (See
Chapter IX.) d. Deposit of gray fissile sandstone which constitutes
the middle formation of the system, characterized also by its peculiar
organic group. (See Chapter VIII.) e. Red and variegated
sandstones, undistinguishable often in their mineral character from
the upper sandstones, c, but in general less gritty, and containing
fewer pebbles, f. Bituminous schists, g. Coarse gritty sandstone. h.
Great Conglomerate. These four beds compose a lower formation of
the system, more strikingly marked by its peculiar organisms than
even the other two. (See Chapters II. III. IV. and V.) In the section
this lower formation is represented as we find it developed in
Caithness and Orkney. In fig. 5 it is represented as developed in
Cromarty, where, though the fossils are identical with those of the
more northern localities, at least one of the deposits, f, is
mineralogically different—alternating beds of sandstone and clay,
these last enclosing limestone nodules, taking the place of the
bituminous schists.

SECTION II.
The Old Red System of England and Wales, as given in the
general Section of Mr. Murchison, with the Silurian Rocks beneath
and the carboniferous limestone above. i. The point in the geological
scale at which vertebrated existences first appear. The three Old Red
Sandstone formations of this section correspond in their
characteristic fossils with those of Scotland, but the proportions in
which they are developed are widely different. The tilestones seem a
comparatively narrow stripe in the system in England; the answering
formation in Scotland, e, f, g, h, is of such enormous thickness, that
it has been held by very superior geologists to contain three distinct
formations—e, the New Red Sandstone, f, a representative of the
Coal Measures, and g, h, the Old Red Sandstone.

SECTION III.
Interesting case of extensive denudation from existing causes on
the northern shore of the Moray Frith. (See pages 197 and 198.) The
figures and letters which mark the various beds correspond with
those of fig. 5, and of the following section. The "fish-bed," No. 1,
represents what the reader will find described in pp. 221-225 as the
"platform of sudden death."

SECTION IV.
Illustration of a fault in the Burn of Eathie, Cromartyshire. (See
pages 204 and 205.)
 EXPLANATIONS OF THE PLATES.
Plate I.—Fig. 1, Restoration of upper side of the elongated
species of Pterichthys (P. oblongus,) referred to in page 47. Fig. 2,
Pterichthys Milleri. Fig. 3, Part of tail of elongated species, showing
portions of the original covering of rhomboidal scales. Fig. 4,
Tubercles of Pterichthys magnified.
Plate II.—Fig. 2, Restoration of under side of Pterichthys
oblongus. Fig. 1, A second specimen of Pterichthys Milleri. Fig. 3,
Portion of wing, natural size.
Plate III.—Fig. 1, Coccosteus cuspidatus. Fig. 2, Impression of
inner surface of large dorsal plate. Fig. 3, Abdominal lozenge-shaped
plate. Fig. 4, Portion of jaw, with teeth.
Plate IV.—Fig. 1, Restoration of Osteolepis major. Fig. 2, Scales
from the upper part of the body magnified. Fig. 3, Large defensive
scale which runs laterally along all the single fins. Fig. 4, Under side
of scale, showing the attaching bar. Fig. 5, Enamelled and
punctulated jaw of the creature. Fig. 6, Magnified portion of fin,
showing the enamelled and punctulated rays.
Plate V.—Fig. 1, Dipterus macrolepidotus. This figure serves
merely to show the place of the fins and the general outline of the
ichthyolite. All the specimens the writer has hitherto examined fail to
show the minuter details. Fig. 2, Glyptolepis leptopterus. Fig. 3,
Single scale of the creature, showing its rustic style of ornament.
Fig. 4, Scale with a nail-like attachment. Fig. 5, Under side of scale.
Fig. 6, Magnified portion of fin. Fig. 7, Shells of the Old Red
Sandstone.
Plate VI.—Fig. 1, Cheirolepis Cummingiæ. Fig. 2, Magnified
scales. Fig. 3, Magnified portion of fin.
Plate VII.—Fig. 1, Cheiracanthus microlepidotus. Fig. 2, Magnified
scales. Figs. 3, 4, 5, 6, 7, 8, Vegetable impressions of the Old Red
Sandstone.
 Plate VIII.—Fig. 1, Diplacanthus longispinus. Fig. 2, Diplacanthus
striatus. Fig. 3, Magnified scales of fig. 1. Fig. 4, Spine of fig. 2,
slightly magnified.
Plate IX.—Fig. 1, One of the tail flaps of the gigantic Crustacean
of Forfarshire. Fig. 2, Reticulated markings of Carmylie.
Plate X.—Fig. 1, Cephalaspis Lyellii, copied from Lyell's Elements
of Geology, Fig. 2, Holoptychius nobilissimus, copied on a greatly
reduced scale from Murchison's Silurian System, Fig. 3, Scale of
Holoptychius, natural size. Fig, 4, Tooth of ditto, also natural size.
These last drawn from specimens in the collection of Mr. Patrick Duff,
of Elgin.

DIRECTIONS TO THE BINDER.

Sheet of Sections to front Title-Page.
Plate I. to front page 44
II. " " " 46
III. " " " 48
IV. " " " 66
V. " " " 72
VI. " " " 78
VII. " " " 82
VIII. " " " 84
IX. " " " 136
X. " " " 154

NEW WALKS IN AN OLD

FIELD;
 OR,
THE OLD RED SANDSTONE.
 CHAPTER I.
The Working-man's True Policy.—His only Mode of acquiring Power.—The Exercise
of the Faculties essential to Enjoyment.—No necessary Connection between
Labor and Unhappiness.—Narrative.—Scenes in a Quarry.—The two dead Birds.
—Landscape.—Ripple Markings on a Sandstone Slab.—Boulder Stones.—
Inference derived from their water-worn Appearance.—Sea-coast Section.—My
first discovered Fossil,—Lias Deposit on the Shores of the Moray Frith.—
Belemnite.—Result of the Experience of half a Lifetime of Toil.—Advantages of
a Wandering Profession in Connection with the Geology of a Country.—
Geological Opportunities of the Stone-Mason.—Design of the present Work.

My advice to young working-men, desirous of bettering their

circumstances, and adding to the amount of their enjoyment, is a
very simple one. Do not seek happiness in what is misnamed
pleasure; seek it rather in what is termed study. Keep your
consciences clear, your curiosity fresh, and embrace every
opportunity of cultivating your minds. You will gain nothing by
attending Chartist meetings. The fellows who speak nonsense with
fluency at these assemblies, and deem their nonsense eloquence,
are totally unable to help either you or themselves; or, if they do
succeed in helping themselves, it will be all at your expense. Leave
them to harangue unheeded, and set yourselves to occupy your
leisure hours in making yourselves wiser men. Learn to make a right
use of your eyes: the commonest things are worth looking at—even
stones and weeds, and the most familiar animals. Head good books,
not forgetting the best of all: there is more true philosophy in the
Bible than in every work of every sceptic that ever wrote; and we
would be all miserable creatures without it, and none more
miserable than you. You are jealous of the upper classes; and
perhaps it is too true that, with some good, you have received much
evil at their hands. It must be confessed they have hitherto been
doing comparatively little for you, and a great deal for themselves.
But upper and lower classes there must be, so long as the world
lasts; and there is only one way in which your jealousy of them can
be well directed. Do not let them get ahead of you in intelligence. It
would be alike unwise and unjust to attempt casting them down to
your own level, and no class would suffer more in the attempt than
yourselves; for you would only be clearing the way, at an immense
expense of blood, and under a tremendous pressure of misery, for
another and perhaps worse aristocracy, with some second Cromwell
or Napoleon at their head. Society, however, is in a state of continual
flux: some in the upper classes are from time to time going down,
and some of you from time to time mounting up to take their places
—always the more steady and intelligent among you, remember;
and if all your minds were cultivated, not merely intellectually, but
morally also, you would find yourselves, as a body, in the possession
of a power which every charter in the world could not confer upon
you, and which all the tyranny or injustice of the world could not
withstand.
I intended, however, to speak rather of the pleasure to be
derived, by even the humblest, in the pursuit of knowledge, than of
the power with which knowledge in the masses is invariably
accompanied. For it is surely of greater importance that men should
receive accessions to their own happiness, than to the influence
which they exert over other men. There is none of the intellectual,
and none of the moral faculties, the exercise of which does not lead
to enjoyment; nay, it is chiefly in the active employment of these
that all enjoyment consists; and hence it is that happiness bears so
little reference to station. It is a truth which has been often told, but
very little heeded or little calculated upon, that though one
nobleman may be happier than another, and one laborer happier
than another, yet it cannot be at all premised of their respective
orders, that the one is in any degree happier than the other. Simple
as the fact may seem, if universally recognized, it would save a great
deal of useless discontent, and a great deal of envy. Will my humbler
readers permit me at once to illustrate this subject, and to introduce
the chapters which follow, by a piece of simple narrative? I wish to
show them how possible it is to enjoy much happiness in very mean
employments. Cowper tells us that labor, though the primal curse,
"has been softened into mercy;" and I think that, even had he not
done so, I would have found out the fact for myself.
It was twenty years, last February, since I set out a little before
sunrise to make my first acquaintance with a life of labor and
restraint, and I have rarely had a heavier heart than on that
morning. I was but a slim, loose-jointed boy at the time—fond of the
pretty intangibilities of romance, and of dreaming when broad
awake; and, woeful change! I was now going to work at what Burns
has instanced in his "Twa Dogs" as one of the most disagreeable of
all employments—to work in a quarry. Bating the passing uneasiness
occasioned by a few gloomy anticipations, the portion of my life
which had already gone by had been happy beyond the common lot.
I had been a wanderer among rocks and woods—a reader of curious
books when I could get them—a gleaner of old traditionary stories;
and now I was going to exchange all my day-dreams, and all my
amusements, for the kind of life in which men toil every day that
they may be enabled to eat, and eat every day that they may be
enabled to toil!
The quarry in which I wrought lay on the southern shore of a
noble inland bay, or frith, rather, with a little clear stream on the one
side, and a thick fir wood on the other. It had been opened in the
Old Red Sandstone of the district, and was overtopped by a huge
bank of diluvial clay, which rose over it in some places to the height
of nearly thirty feet, and which at this time was rent and shivered,
wherever it presented an open front to the weather, by a recent
frost. A heap of loose fragments, which had fallen from above,
blocked up the face of the quarry, and my first employment was to
clear them away. The friction of the shovel soon blistered my hands;
but the pain was by no means very severe, and I wrought hard and
willingly, that I might see how the huge strata below, which
presented so firm and unbroken a frontage, were to be torn up and
removed. Picks, and wedges, and levers were applied by my brother-
workmen; and simple and rude as I had been accustomed to regard
these implements, I found I had much to learn in the way of using
them. They all proved inefficient, however; and the workmen had to
bore into one of the inferior strata, and employ gunpowder. The
process was new to me, and I deemed it a highly amusing one: it
had the merit, too, of being attended with some such degree of
danger as a boating or rock excursion, and had thus an interest
independent of its novelty. We had a few capital shots: the
fragments flew in every direction; and an immense mass of the
diluvium came toppling down, bearing with it two dead birds, that in
a recent storm had crept into one of the deeper fissures, to die in
the shelter. I felt a new interest in examining them. The one was a
pretty cock goldfinch, with its hood of vermilion, and its wings inlaid
with the gold to which it owes its name, as unsoiled and smooth as
if it had been preserved for a museum. The other, a somewhat rarer
bird, of the woodpecker tribe, was variegated with light blue and a
grayish yellow. I was engaged in admiring the poor little things,
more disposed to be sentimental, perhaps, than if I had been ten
years older, and thinking of the contrast between the warmth and
jollity of their green summer haunts, and the cold and darkness of
their last retreat, when I heard our employer bidding the workmen
lay by their tools. I looked up, and saw the sun sinking behind the
thick fir wood beside us, and the long, dark shadows of the trees
stretching downwards towards the shore.
This was no very formidable beginning of the course of life I had
so much dreaded. To be sure, my hands were a little sore, and I felt
nearly as much fatigued as if I had been climbing among the rocks;
but I had wrought and been useful, and had yet enjoyed the day
fully as much as usual. It was no small matter, too, that the evening,
converted, by a rare transmutation, into the delicious "blink of rest"
which Burns so truthfully describes, was all my own. I was as light of
heart next morning as any of my brother-workmen. There had been
a smart frost during the night, and the rime lay white on the grass
as we passed onwards through the fields; but the sun rose in a clear
atmosphere, and the day mellowed, as it advanced, into one of
those delightful days of early spring, which give so pleasing an
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookname.com