Explore the full ebook collection and download it now at textbookfull.

com

Theory of stochastic objects: probability,

stochastic processes, and inference First Edition
Micheas

https://textbookfull.com/product/theory-of-stochastic-
objects-probability-stochastic-processes-and-inference-
first-edition-micheas/

OR CLICK HERE

DOWLOAD EBOOK

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

Click here to visit textbookfull.com and download textbook now
 Your digital treasures (PDF, ePub, MOBI) await
Download instantly and pick your perfect format...

Read anywhere, anytime, on any device!

Theory of stochastic objects probability stochastic

processes and inference 1st Edition Micheas

https://textbookfull.com/product/theory-of-stochastic-objects-
probability-stochastic-processes-and-inference-1st-edition-micheas/

textbookfull.com

Biota Grow 2C gather 2C cook Loucas

https://textbookfull.com/product/biota-grow-2c-gather-2c-cook-loucas/

textbookfull.com

Applied Probability and Stochastic Processes Second

Edition Beichelt

https://textbookfull.com/product/applied-probability-and-stochastic-
processes-second-edition-beichelt/

textbookfull.com

Applied Probability and Stochastic Processes Second

Edition Beichelt

https://textbookfull.com/product/applied-probability-and-stochastic-
processes-second-edition-beichelt-2/

textbookfull.com
Fundamentals of probability: with stochastic processes
Fourth Edition Ghahramani

https://textbookfull.com/product/fundamentals-of-probability-with-
stochastic-processes-fourth-edition-ghahramani/

textbookfull.com

Generalized Statistical Thermodynamics Thermodynamics of

Probability Distributions and Stochastic Processes Themis
Matsoukas
https://textbookfull.com/product/generalized-statistical-
thermodynamics-thermodynamics-of-probability-distributions-and-
stochastic-processes-themis-matsoukas/
textbookfull.com

Applied Probability From Random Sequences to Stochastic

Processes Valérie Girardin

https://textbookfull.com/product/applied-probability-from-random-
sequences-to-stochastic-processes-valerie-girardin/

textbookfull.com

Fundamentals of Probability With Stochastic Processes

Solution Manual 4th Edition Ghahramani Saeed

https://textbookfull.com/product/fundamentals-of-probability-with-
stochastic-processes-solution-manual-4th-edition-ghahramani-saeed/

textbookfull.com

Theory and Statistical Applications of Stochastic

Processes 1st Edition Yuliya Mishura

https://textbookfull.com/product/theory-and-statistical-applications-
of-stochastic-processes-1st-edition-yuliya-mishura/

textbookfull.com
 Theory of
Stochastic Objects
Probability, Stochastic Processes
and Inference
CHAPMAN & HALL/CRC
Texts in Statistical Science Series
Series Editors
Joseph K. Blitzstein, Harvard University, USA
Julian J. Faraway, University of Bath, UK
Martin Tanner, Northwestern University, USA
Jim Zidek, University of British Columbia, Canada

Statistical Theory: A Concise Introduction Analysis of Variance, Design, and Regression:

F. Abramovich and Y. Ritov Linear Modeling for Unbalanced Data,
Practical Multivariate Analysis, Fifth Edition Second Edition
A. Afifi, S. May, and V.A. Clark R. Christensen
Practical Statistics for Medical Research Bayesian Ideas and Data Analysis: An Introduction
D.G. Altman for Scientists and Statisticians
R. Christensen, W. Johnson, A. Branscum,
Interpreting Data: A First Course
and T.E. Hanson
in Statistics
A.J.B. Anderson Modelling Binary Data, Second Edition
D. Collett
Introduction to Probability with R
K. Baclawski Modelling Survival Data in Medical Research, Third
Edition
Linear Algebra and Matrix Analysis for Statistics
D. Collett
S. Banerjee and A. Roy
Introduction to Statistical Methods for
Modern Data Science with R
Clinical Trials
B. S. Baumer, D. T. Kaplan, and N. J. Horton
T.D. Cook and D.L. DeMets
Mathematical Statistics: Basic Ideas and Selected
Applied Statistics: Principles and Examples
Topics, Volume I, Second Edition
P. J. Bickel and K. A. Doksum D.R. Cox and E.J. Snell

Mathematical Statistics: Basic Ideas and Selected Multivariate Survival Analysis and Competing Risks
Topics, Volume II M. Crowder
P. J. Bickel and K. A. Doksum Statistical Analysis of Reliability Data
Analysis of Categorical Data with R M.J. Crowder, A.C. Kimber, T.J. Sweeting,
C. R. Bilder and T. M. Loughin and R.L. Smith

Statistical Methods for SPC and TQM An Introduction to Generalized Linear Models,
D. Bissell Third Edition
A.J. Dobson and A.G. Barnett
Introduction to Probability
J. K. Blitzstein and J. Hwang Nonlinear Time Series: Theory, Methods, and
Applications with R Examples
Bayesian Methods for Data Analysis, Third Edition R. Douc, E. Moulines, and D.S. Stoffer
B.P. Carlin and T.A. Louis
Introduction to Optimization Methods and Their
Statistics in Research and Development, Applications in Statistics
Second Edition B.S. Everitt
R. Caulcutt
Extending the Linear Model with R: Generalized
The Analysis of Time Series: An Introduction, Linear, Mixed Effects and Nonparametric Regression
Sixth Edition Models, Second Edition
C. Chatfield J.J. Faraway
Introduction to Multivariate Analysis Linear Models with R, Second Edition
C. Chatfield and A.J. Collins J.J. Faraway
Problem Solving: A Statistician’s Guide, A Course in Large Sample Theory
Second Edition T.S. Ferguson
C. Chatfield
Multivariate Statistics: A Practical Approach
Statistics for Technology: A Course in Applied B. Flury and H. Riedwyl
Statistics, Third Edition
C. Chatfield
Readings in Decision Analysis Exercises and Solutions in Statistical Theory
S. French L.L. Kupper, B.H. Neelon, and S.M. O’Brien
Discrete Data Analysis with R: Visualization and Design and Analysis of Experiments with R
Modeling Techniques for Categorical and Count J. Lawson
Data Design and Analysis of Experiments with SAS
M. Friendly and D. Meyer J. Lawson
Markov Chain Monte Carlo: Stochastic Simulation A Course in Categorical Data Analysis
for Bayesian Inference, Second Edition T. Leonard
D. Gamerman and H.F. Lopes
Statistics for Accountants
Bayesian Data Analysis, Third Edition S. Letchford
A. Gelman, J.B. Carlin, H.S. Stern, D.B. Dunson,
A. Vehtari, and D.B. Rubin Introduction to the Theory of Statistical Inference
H. Liero and S. Zwanzig
Multivariate Analysis of Variance and Repeated
Measures: A Practical Approach for Behavioural Statistical Theory, Fourth Edition
Scientists B.W. Lindgren
D.J. Hand and C.C. Taylor Stationary Stochastic Processes: Theory and
Practical Longitudinal Data Analysis Applications
D.J. Hand and M. Crowder G. Lindgren

Linear Models and the Relevant Distributions and Statistics for Finance
Matrix Algebra E. Lindström, H. Madsen, and J. N. Nielsen
D.A. Harville The BUGS Book: A Practical Introduction to
Logistic Regression Models Bayesian Analysis
J.M. Hilbe D. Lunn, C. Jackson, N. Best, A. Thomas, and
D. Spiegelhalter
Richly Parameterized Linear Models: Additive, Time
Series, and Spatial Models Using Random Effects Introduction to General and Generalized
J.S. Hodges Linear Models
H. Madsen and P. Thyregod
Statistics for Epidemiology
N.P. Jewell Time Series Analysis
H. Madsen
Stochastic Processes: An Introduction, Third Edition
P.W. Jones and P. Smith Pólya Urn Models
H. Mahmoud
The Theory of Linear Models
B. Jørgensen Randomization, Bootstrap and Monte Carlo
Methods in Biology, Third Edition
Pragmatics of Uncertainty B.F.J. Manly
J.B. Kadane
Statistical Regression and Classification:
Principles of Uncertainty From Linear Models to Machine Learning
J.B. Kadane N. Matloff
Graphics for Statistics and Data Analysis with R Introduction to Randomized Controlled Clinical
K.J. Keen Trials, Second Edition
Mathematical Statistics J.N.S. Matthews
K. Knight Statistical Rethinking: A Bayesian Course with
Introduction to Functional Data Analysis Examples in R and Stan
P. Kokoszka and M. Reimherr R. McElreath
Introduction to Multivariate Analysis: Linear and Statistical Methods in Agriculture and Experimental
Nonlinear Modeling Biology, Second Edition
S. Konishi R. Mead, R.N. Curnow, and A.M. Hasted
Nonparametric Methods in Statistics with SAS Statistics in Engineering: A Practical Approach
Applications A.V. Metcalfe
O. Korosteleva Theory of Stochastic Objects: Probability, Stochastic
Modeling and Analysis of Stochastic Systems, Processes and Inference
Third Edition A.C. Micheas
V.G. Kulkarni Statistical Inference: An Integrated Approach,
Exercises and Solutions in Biostatistical Theory Second Edition
L.L. Kupper, B.H. Neelon, and S.M. O’Brien H. S. Migon, D. Gamerman, and F. Louzada
Beyond ANOVA: Basics of Applied Statistics Spatio-Temporal Methods in Environmental
R.G. Miller, Jr. Epidemiology
G. Shaddick and J.V. Zidek
A Primer on Linear Models
J.F. Monahan Decision Analysis: A Bayesian Approach
J.Q. Smith
Stochastic Processes: From Applications to Theory
P.D Moral and S. Penev Analysis of Failure and Survival Data
P. J. Smith
Applied Stochastic Modelling, Second Edition
B.J.T. Morgan Applied Statistics: Handbook of GENSTAT
Analyses
Elements of Simulation
B.J.T. Morgan E.J. Snell and H. Simpson

Probability: Methods and Measurement Applied Nonparametric Statistical Methods,

A. O’Hagan Fourth Edition
P. Sprent and N.C. Smeeton
Introduction to Statistical Limit Theory
A.M. Polansky Data Driven Statistical Methods
P. Sprent
Applied Bayesian Forecasting and Time Series
Analysis Generalized Linear Mixed Models:
A. Pole, M. West, and J. Harrison Modern Concepts, Methods and Applications
W. W. Stroup
Statistics in Research and Development,
Time Series: Modeling, Computation, and Inference Survival Analysis Using S: Analysis of
R. Prado and M. West Time-to-Event Data
M. Tableman and J.S. Kim
Essentials of Probability Theory for Statisticians
M.A. Proschan and P.A. Shaw Applied Categorical and Count Data Analysis
W. Tang, H. He, and X.M. Tu
Introduction to Statistical Process Control
P. Qiu Elementary Applications of Probability Theory,
Second Edition
Sampling Methodologies with Applications H.C. Tuckwell
P.S.R.S. Rao
Introduction to Statistical Inference and Its
A First Course in Linear Model Theory Applications with R
N. Ravishanker and D.K. Dey M.W. Trosset
Essential Statistics, Fourth Edition Understanding Advanced Statistical Methods
D.A.G. Rees P.H. Westfall and K.S.S. Henning
Stochastic Modeling and Mathematical Statistics: Statistical Process Control: Theory and Practice,
A Text for Statisticians and Quantitative Scientists Third Edition
F.J. Samaniego G.B. Wetherill and D.W. Brown
Statistical Methods for Spatial Data Analysis Generalized Additive Models: An Introduction
O. Schabenberger and C.A. Gotway with R, Second Edition
Bayesian Networks: With Examples in R S. Wood
M. Scutari and J.-B. Denis Epidemiology: Study Design and
Large Sample Methods in Statistics Data Analysis, Third Edition
P.K. Sen and J. da Motta Singer M. Woodward
Introduction to Statistical Methods for Financial Practical Data Analysis for Designed Experiments
Models B.S. Yandell
T. A. Severini
 Texts in Statistical Science

Theory of
Stochastic Objects
Probability, Stochastic Processes
and Inference

Athanasios Christou Micheas

Department of Statistics, University of Missouri, USA
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2018 by Taylor & Francis Group, LLC
CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works
Printed on acid-free paper
Version Date: 20171219
International Standard Book Number-13: 978-1-4665-1520-8 (Hardback)
This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been
made to publish reliable data and information, but the author and publisher cannot assume responsibility for the
validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the
copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to
publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let
us know so we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or
utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including
photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission
from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copyright.com
(http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA
01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For
organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for
identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data

Names: Micheas, Athanasios Christou, author.

Title: Theory of stochastic objects : probability, stochastic processes and
inference / by Athanasios Christou Micheas.
Description: Boca Raton, Florida : CRC Press, [2018] | Includes
bibliographical references and index.
Identifiers: LCCN 2017043053| ISBN 9781466515208 (hardback) | ISBN
9781315156705 (e-book)
Subjects: LCSH: Point processes. | Stochastic processes.
Classification: LCC QA274.42 .M53 2018 | DDC 519.2/3--dc23
LC record available at https://lccn.loc.gov/2017043053

Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
To my family
 Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi

List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii

List of Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . xxv

List of Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxvii

List of Distribution Notations . . . . . . . . . . . . . . . . . . . . . . xxix

1 Rudimentary Models and Simulation Methods . . . . . . . . . . . 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Rudimentary Probability . . . . . . . . . . . . . . . . . . . . . . 2
1.2.1 Probability Distributions . . . . . . . . . . . . . . . . . . . 4
1.2.2 Expectation . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.2.3 Mixtures of Distributions . . . . . . . . . . . . . . . . . . 9
1.2.4 Transformations of Random Vectors . . . . . . . . . . . . . 11
1.3 The Bayesian Approach . . . . . . . . . . . . . . . . . . . . . . 11
1.3.1 Conjugacy . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.3.2 General Prior Selection Methods and Properties . . . . . . 14
1.3.3 Hierarchical Bayesian Models . . . . . . . . . . . . . . . . 15
1.4 Simulation Methods . . . . . . . . . . . . . . . . . . . . . . . . 17
1.4.1 The Inverse Transform Algorithm . . . . . . . . . . . . . . 17
1.4.2 The Acceptance-Rejection Algorithm . . . . . . . . . . . . 18
1.4.3 The Composition Method for Generating Mixtures . . . . . 20
1.4.4 Generating Multivariate Normal and Related Distributions . 20
1.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

2 Statistical Inference . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2 Decision Theory . . . . . . . . . . . . . . . . . . . . . . . . . . 27
ix
x CONTENTS
2.3 Point Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.3.1 Classical Methods . . . . . . . . . . . . . . . . . . . . . . 30
2.3.2 Bayesian Approach . . . . . . . . . . . . . . . . . . . . . . 39
2.3.3 Evaluating Point Estimators Using Decision Theory . . . . 40
2.3.4 Convergence Concepts and Asymptotic Behavior . . . . . . 42
2.4 Interval Estimation . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.4.1 Confidence Intervals . . . . . . . . . . . . . . . . . . . . . 45
2.4.2 Highest Posterior Density Credible Sets . . . . . . . . . . . 46
2.4.3 Decision Theoretic . . . . . . . . . . . . . . . . . . . . . . 48
2.5 Hypothesis Testing . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.5.1 Classic Methods . . . . . . . . . . . . . . . . . . . . . . . 49
2.5.2 Bayesian Testing Procedures . . . . . . . . . . . . . . . . . 55
2.5.3 Decision Theoretic . . . . . . . . . . . . . . . . . . . . . . 58
2.5.4 Classical and Bayesian p-values . . . . . . . . . . . . . . . 60
2.5.5 Reconciling the Bayesian and Classical Paradigms . . . . . 65
2.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
2.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

3 Measure and Integration Theory . . . . . . . . . . . . . . . . . . . 77

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.2 Deterministic Set Theory . . . . . . . . . . . . . . . . . . . . . . 77
3.3 Topological Spaces and σ-fields . . . . . . . . . . . . . . . . . . 79
3.4 Product Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
3.5 Measurable Spaces and Mappings . . . . . . . . . . . . . . . . . 86
3.6 Measure Theory and Measure Spaces . . . . . . . . . . . . . . . 91
3.6.1 Signed Measures and Decomposition Theorems . . . . . . 97
3.6.2 Carathéodory Measurability and Extension Theorem . . . . 100
3.6.3 Construction of the Lebesgue Measure . . . . . . . . . . . 103
3.7 Defining Integrals with Respect to Measures . . . . . . . . . . . . 106
3.7.1 Change of Variable and Integration over Sets . . . . . . . . 113
3.7.2 Lebesgue, Riemann and Riemann-Stieltjes Integrals . . . . 115
3.7.3 Radon-Nikodym Theorem . . . . . . . . . . . . . . . . . . 118
3.7.4 Product Measure and Fubini Theorem . . . . . . . . . . . . 121
3.7.5 L p -spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
3.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
3.9 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

4 Probability Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
4.2 Probability Measures and Probability Spaces . . . . . . . . . . . 139
4.2.1 Extension of Probability Measure . . . . . . . . . . . . . . 140
4.2.2 Defining Random Objects . . . . . . . . . . . . . . . . . . 143
4.2.3 Distribution Functions and Densities . . . . . . . . . . . . 148
CONTENTS xi
4.2.4 Independence . . . . . . . . . . . . . . . . . . . . . . . . . 153
4.2.5 Calculating Probabilities for Limits of Events . . . . . . . . 158
4.2.6 Expectation of a Random Object . . . . . . . . . . . . . . . 159
4.2.7 Characteristic Functions . . . . . . . . . . . . . . . . . . . 162
4.3 Conditional Probability . . . . . . . . . . . . . . . . . . . . . . . 165
4.3.1 Conditioning on the Value of a Random Variable . . . . . . 165
4.3.2 Conditional Probability and Expectation Given a σ-field . . 170
4.3.3 Conditional Independence Given a σ-field . . . . . . . . . 177
4.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
4.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179

5 Convergence of Random Objects . . . . . . . . . . . . . . . . . . 183

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
5.2 Existence of Independent Sequences of Random Variables . . . . 183
5.3 Limiting Behavior of Sequences of Random Variables . . . . . . 184
5.3.1 Slutsky and Cramér Theorems . . . . . . . . . . . . . . . . 184
5.3.2 Consistency of the MLE . . . . . . . . . . . . . . . . . . . 186
5.4 Limiting Behavior of Probability Measures . . . . . . . . . . . . 187
5.4.1 Integrating Probability Measures to the Limit . . . . . . . . 187
5.4.2 Compactness of the Space of Distribution Functions . . . . 191
5.4.3 Weak Convergence via Non-Central Moments . . . . . . . 193
5.5 Random Series . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
5.5.1 Convolutions . . . . . . . . . . . . . . . . . . . . . . . . . 194
5.5.2 Fourier Inversion and the Continuity Theorem . . . . . . . 196
5.5.3 Limiting Behavior of Partial Sums . . . . . . . . . . . . . . 199
5.5.4 Central Limit Theorems . . . . . . . . . . . . . . . . . . . 200
5.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
5.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204

6 Random Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . 207

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
6.2 Definitions and Properties . . . . . . . . . . . . . . . . . . . . . 207
6.3 Martingales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
6.3.1 Filtrations and Stopping Times . . . . . . . . . . . . . . . 215
6.3.2 Convergence Theorems . . . . . . . . . . . . . . . . . . . 216
6.3.3 Applications . . . . . . . . . . . . . . . . . . . . . . . . . 223
6.3.4 Wald Identities and Random Walks . . . . . . . . . . . . . 225
6.4 Renewal Sequences . . . . . . . . . . . . . . . . . . . . . . . . . 226
6.5 Markov Chains . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
6.5.1 Definitions and Properties . . . . . . . . . . . . . . . . . . 233
6.5.2 Discrete State Space . . . . . . . . . . . . . . . . . . . . . 236
6.5.3 The Martingale Problem . . . . . . . . . . . . . . . . . . . 240
6.5.4 Visits to Fixed States: General State Space . . . . . . . . . 241
xii CONTENTS
6.5.5 Visits to Fixed States: Discrete State Space . . . . . . . . . 243
6.5.6 State Classification . . . . . . . . . . . . . . . . . . . . . . 248
6.5.7 Limiting Behavior . . . . . . . . . . . . . . . . . . . . . . 255
6.6 Stationary Sequences and Ergodicity . . . . . . . . . . . . . . . . 258
6.7 Applications to Markov Chain Monte Carlo Methods . . . . . . . 262
6.7.1 Metropolis-Hastings Algorithm . . . . . . . . . . . . . . . 263
6.7.2 Gibbs Sampling . . . . . . . . . . . . . . . . . . . . . . . 264
6.7.3 Reversible Jump Markov Chain Monte Carlo . . . . . . . . 265
6.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
6.9 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269

7 Stochastic Processes . . . . . . . . . . . . . . . . . . . . . . . . . 275

7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
7.2 The Poisson Process . . . . . . . . . . . . . . . . . . . . . . . . 277
7.3 General Stochastic Processes . . . . . . . . . . . . . . . . . . . . 279
7.3.1 Continuous Time Filtrations and Stopping Times . . . . . . 281
7.3.2 Continuous Time Martingales . . . . . . . . . . . . . . . . 283
7.3.3 Kolmogorov Existence Theorem . . . . . . . . . . . . . . . 285
7.4 Markov Processes . . . . . . . . . . . . . . . . . . . . . . . . . . 289
7.4.1 Càdlàg Space . . . . . . . . . . . . . . . . . . . . . . . . . 289
7.4.2 Infinitesimal Generators . . . . . . . . . . . . . . . . . . . 294
7.4.3 The Martingale Problem . . . . . . . . . . . . . . . . . . . 296
7.4.4 Construction via Subordinated Markov Chains . . . . . . . 297
7.4.5 Discrete State Space . . . . . . . . . . . . . . . . . . . . . 298
7.4.5.1 Sample Paths and State Classification . . . . . . . . 301
7.4.5.2 Construction via Jump Times . . . . . . . . . . . . 303
7.4.5.3 Infinitesimal Generator and Transition Function . . 305
7.4.5.4 Limiting Behavior . . . . . . . . . . . . . . . . . . 307
7.4.5.5 Birth-Death Processes . . . . . . . . . . . . . . . . 309
7.4.6 Brownian Motion . . . . . . . . . . . . . . . . . . . . . . . 312
7.4.7 Construction of the Wiener Measure . . . . . . . . . . . . . 314
7.5 Building on Martingales and Brownian Motion . . . . . . . . . . 317
7.5.1 Stochastic Integrals . . . . . . . . . . . . . . . . . . . . . . 317
7.5.2 Stochastic Differential Equations . . . . . . . . . . . . . . 324
7.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
7.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332

A Additional Topics and Complements . . . . . . . . . . . . . . . . 337

p
A.1 Mappings in R . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
A.2 Topological, Measurable and Metric Spaces . . . . . . . . . . . . 338
A.3 Baire Functions and Spaces . . . . . . . . . . . . . . . . . . . . . 342
A.4 Fisher Information . . . . . . . . . . . . . . . . . . . . . . . . . 343
A.5 Multivariate Analysis Topics . . . . . . . . . . . . . . . . . . . . 344
CONTENTS xiii
A.6 State Classification . . . . . . . . . . . . . . . . . . . . . . . . . 345
A.7 MATLAB R Code Function Calls . . . . . . . . . . . . . . . . . 346
A.8 Commonly Used Distributions and Their Densities . . . . . . . . 347

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
 Preface

Random variables and random vectors have been well defined and studied for
over a century. Subsequently, in the history of statistical science, researchers began
considering collections of points together, which gave birth to point process theory
and more recently, to random set theory. This was mainly motivated due to advances
in technology and the types of data that experimenters began investigating, which
in turn led to the creation and investigation of advanced statistical methods able to
handle such data.
In this book we take the reader on a journey through some of the most essential
topics in mathematics and statistics, constantly building on previous concepts, mak-
ing the transition from elementary statistical inference to the advanced probabilistic
treatment more natural and concrete. Our central focus is defining and exploring the
concept of a random quantity or object in different contexts, where depending on
the data under consideration, “random objects” are described using random vari-
ables, vectors or matrices, stochastic processes, integrals and differential equations,
or point processes and random sets.
This view of random objects has not been adequately investigated and pre-
sented in mathematics and statistics textbooks that are out there since they have
mostly concentrated on specific parts of the aforementioned concepts. This is one
of the reasons why I undertake the task of writing a textbook that would present
the knowledge in a concrete way, through examples and exercises, which is sorely
needed in understanding statistical inference, probability theory and stochastic pro-
cesses. This approach will help the instructor of these topics to engage the students
through problem sets and present the theory and applications involved in a way that
they will appreciate.
Since this monumental task cannot be accomplished in a single textbook, the
theoretical and modeling topics considered have been organized in two texts; this
text is concerned with rudimentary to advanced theoretical aspects of random ob-
jects based on random variables, including statistical inference, probability theory
and stochastic processes. The modeling of these objects and their applications to
real life data is presented in the text Theory and Modeling of Stochastic Objects:
Point Processes and Random Sets (forthcoming, hereafter referred to as TMSO-
PPRS). The latter stochastic objects are a natural extension of random variables

xv
xvi PREFACE
and vectors and we can think of the TMSO-PPRS text as a natural continuation of
the theory presented herein.
In particular, we present a comprehensive account of topics in statistics in a
way that can be a natural extension of a more traditional graduate course in prob-
ability theory. This is especially true for Chapters 1 and 2, which is a feature that
has been lacking from available texts in probability theory. Another distinguishing
feature of this text is that we have included an amazing amount of material. More
precisely, one would need to use at least one book on real analysis, one in measure
and/or probability theory, one in stochastic processes, and at least one on statistics
to capture just the expository material that has gone into this text.
Being a teacher and mentor to undergraduate and graduate students, I have seen
their attempts to comprehend new material from rudimentary to advanced mathe-
matical and statistical concepts. I have also witnessed their struggles with essential
topics in statistics, such as defining a probability space for a random variable, which
is one of the most important constructs in statistics. This book attempts to introduce
these concepts in a novel way making it more accessible to students and researchers
through examples. This approach is lacking in most textbooks/monographs that one
can use to teach students.
Instructors and researchers in academia often find themselves complementing
material from several books in order to provide a spherical overview of the topics
of a class. This book is the result of my efforts over the years to provide comprehen-
sive and compact accounts of topics I had to teach to undergraduate and graduate
students.
Therefore, the book is targeted toward students at the master’s and Ph.D. levels,
as well as academicians in the mathematics and statistics disciplines. Although the
concepts will be built from the master’s level up, the book addresses advanced read-
ers in the later chapters. When used as a textbook, prior knowledge of probability
or measure theory is welcomed but not necessary.
In particular, Chapters 1 and 2 can be used for several courses on statistical
inference with minor additions for any proofs the instructor chooses to further il-
lustrate. In these chapters we summarize over a century and a half of development
in mathematical statistics. Depending on the level of the course, the instructor can
select specific exercises to supplement the text, in order to provide a better under-
standing and more depth into the concepts under consideration. For example, using
selectively the material and exercises from Chapters 1, 2, 4 and 5, I have taught
several sequences on statistical inference at the University of Missouri (MU), in-
cluding Stat7750/60 and 4750/60 (statistical inference course at the undergraduate
and master’s level), Stat8710/20 (intermediate statistical inference course at the
Ph.D. level) and Stat9710/20 (advanced inference for Ph.D. students).
At the master’s level, it is recommended that the instructor omits advanced top-
ics from Chapter 2, including most of the decision-theoretic topics and the corre-
PREFACE xvii
sponding proofs of the relevant results. Basic theorems and their proofs, such as
the Bayes or the factorization theorem, should be presented to the students in de-
tail. The proofs of such results are included as exercises, and the instructor can use
the solution manual in order to choose what they deem appropriate to illustrate to
the students.
For example, when teaching a statistical inference course for Ph.D. students, all
concepts presented in Chapters 1 and 2 should be introduced, as well as topics on
asymptotics from Chapter 5. However, certain proofs might be beyond the level of
an intermediate statistical inference course for Ph.D. students. For example, when
it comes to introducing evaluation of point estimators, we may omit the explicit
proof of all parts of the important remark 2.12 and simply present the material, or
the compactness results in Chapter 5, and focus only on the central limit theorems
or Slutsky and Cramér theorems.
For an advanced course on statistical inference at the Ph.D. level, one would
omit most of the rudimentary results of Chapter 1, and focus on topics from Chap-
ter 2 (inference), Chapter 4 (e.g., characteristic functions), and Chapter 5 (asymp-
totics), including all the important proofs of the theorems and remarks presented in
the text. Once again, the instructor can find the solution manual invaluable in this
case, since it will allow them to select the topics they want to present along with
concrete proofs.
Chapters 3-5 can be used to introduce measure theoretic probability to mathe-
matics and statistics graduate students. Some of the proofs should be skipped since
it would take more than one semester to go through all the material. More precisely,
over the past decade when I taught the advanced probability theory course Stat9810
at MU, I had to omit most of the measure theoretic proofs and be quite selective in
the material for a one-semester course. For example, important theorems and their
proofs, like Fubini, Kolmogorov 0-1 Law, Radon-Nikodym or Kolmogorov Three
Series, should be illustrated to the students in detail.
In contrast, one may skip the proofs of the theoretical development of the
Carathodory extension theorem, or omit the proofs of the decomposition theorems
(Chapter 3) and the compactness theorems of Chapter 5. Of course, most of the
important results in measure and probability theory and their proofs are still there
for the inquisitive student and researcher who needs to go deeper. These chapters
are fairly comprehensive and self-contained, which is important for Ph.D. students
that have not had an advanced real analysis course.
Chapter 6 is a fairly comprehensive account of stochastic processes in discrete
time and in particular Markov chains. This material has been used to teach an intro-
ductory course on stochastic processes to both undergraduate and master’s students
(Stat4850/7850), as well as Ph.D.-level students in one semester (Stat 9820, a con-
tinuation of Stat9810). Note that most of the development and exposition of discrete
Markov chains and processes does not require heavy measure theory as presented
xviii PREFACE
in Chapters 6 and 7, therefore making it accessible to a wide variety of students, in-
cluding undergraduates. A good working knowledge of matrix algebra is required
in this case, which is a requirement for the undergraduate and graduate students
when they take this course. In particular, the instructor simply needs to explain in a
rudimentary way “transition probability measures,” e.g., replace it with the notion
of transition probabilities and matrices, and then the material can be presented to
the students in a non-measure theoretic way.
The material in Chapter 7 has been used to teach stochastic processes in con-
tinuous time to Ph.D. (Stat 9820) and advanced master’s level students, including
topics from Chapter 6, as mentioned above. The instructor can supplement materi-
als from other chapters as they see fit in order to build the mathematical foundations
of the concepts presented as needed. For example, in the beginning of the class we
may conduct a mini review of probability theory and Markov chains before jumping
into continuous time stochastic processes.
As you begin reading, several features that help with the learning process should
immediately draw your attention; each chapter begins with basic illustrations and
ends with a more advanced treatment of the topic at hand. We are exploring and
reconciling, when feasible, both the frequentist and Bayesian approaches to the
topics considered. In addition, recent developments in statistics are presented or
referenced in the text and summary of each chapter.
Proofs for most of the theorems, lemmas and remarks presented in each chapter
are given in the text or are requested as exercises, with the exception of the rudi-
mentary Chapters 1 and 2, where the proofs are requested as exercises only. Proofs
and additional information on the topics discussed can be found in the books or
journal papers referenced at the summary section of each chapter. Of course, the
interested reader can find proofs to selected exercises in the supplementary online
material for the book (see website below).
The theorems and results presented in the text can range from easy to compli-
cated, and therefore, we usually follow them with an illustrative remark or example
to explain the new concept. To further help in our understanding of the material
and for quick reference, various topics and complements from mathematics and
statistics are included in an appendix.
The MATLAB R code used for the examples presented along with solutions
to exercises and other material, such as errata, can be found at the book website
https://www.crcpress.com/9781466515208.

There are many people that have contributed, in their own way, to the creation
of this book. I am grateful to the faculty members of the Department of Statistics at
the University of Missouri, USA, for their constructive interactions and discussions
over the years. In particular, special thanks go to my friends and colleagues Christo-
pher Wikle, Scott Holan, Stamatis Dostoglou and Joe Cavanaugh (University of
PREFACE xix
Iowa), and my friend and mentor Konstantinos Zografos from the Department of
Mathematics, University of Ioannina, Greece. Lastly, my academic advisor, Distin-
guished Professor of Statistics Dipak Dey, Department of Statistics, University of
Connecticut, USA, has been an inspiration to me over the years.
I am grateful to Professors Stamatis Dostoglou, Department of Mathematics,
University of Missouri, USA, Georg Lindgren, Department of Mathematical Statis-
tics, Centre for Mathematical Sciences, Lund, Sweden, and an anonymous re-
viewer, for their invaluable comments and suggestions regarding earlier versions
of the manuscript. Special thanks go to my friend and colleague Distinguished Pro-
fessor Noel Cressie, School of Mathematics and Applied Statistics, University of
Wollongong, Australia, for his support and encouragement over the years as I was
working on the manuscript, as well as for his advice regarding all aspects of the
book, including its title.
Additional thanks go to the hundreds of students for their undivided attention
while they had to take classes from me on these topics and have helped me better
myself through the teaching process. In particular, special thanks goes to all my
graduate students, especially to Jiaxun Chen and Alex Oard. I am also grateful to
Rob Calver, Michele Dimont, Becky Condit and the friends at Chapman-Hall/CRC
for their patience while the manuscript was composed and for their help with the
copy edit process.
Above all, my appreciation and love to my family, my daughters Vaso, Evi and
Christina, my wife Lada and my father Christos, for their unconditional love and
understanding.
I apologize in advance for any typos or errors in the text and I would be grateful
for any comments, suggestions or corrections the kind reader would like to bring to
my attention.
Sakis Micheas
December 2017
 List of Figures

4.1 Displaying random objects: (a) random functions, (b)-(d) random

counting measures showing regularity, randomness and clustering
of the points, (e) random discs, and (f) Gaussian random field. . . 149

7.1 Brownian motion realizations: (a) univariate, (b) bivariate. . . . . 314

xxi
 List of Tables

2.1 Schematic for any hypothesis testing problem along with the
occurrence of the Type I and II errors. . . . . . . . . . . . . . . . 50
2.2 Simulations of Monte Carlo goodness-of-fit tests. We used
L = 100000 predictive samples for n = 10, 50, 100 observed sample
points and the data is simulated from three models Uni f (0, 1),
Gamma(10, 10), and N(−10, 1). We choose λ = 1 for the
hyperparameter, and p pred is provided for four statistics
T 1 (X) = X(1) , T 2 (X) = X(n) , T 3 (X) = X, and T 4 (X) = S 2 . Based on
these results T 1 emerges as the best test statistic in order to assess
the entertained model. . . . . . . . . . . . . . . . . . . . . . . . 64

4.1 Characteristic functions of commonly used continuous

distributions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
4.2 Characteristic functions of commonly used discrete distributions. 166

xxiii
 List of Abbreviations

a.s. Almost Surely

BCT Bounded Convergence Theorem
Be Beta function
Càdlàg Continue à Droite, Limites à Gauche (RCLL)
cdf Cumulative Distribution Function
cf Characteristic Function
CI Confidence Interval
CLT Central Limit Theorem
CR-LB Cramer-Rao Lower Bound
DCT Dominated Convergence Theorem
ev. Eventually
HBM Hierarchical Bayesian Model
HPD Highest Posterior Density
iid Independent and Identically Distributed
i.o. Infinitely Often
LHS Left-Hand Side
MAP Maximum a Posteriori
MCMC Markov Chain Monte Carlo
MCT Monotone Convergence Theorem
MLE Maximum Likelihood Estimator
MG Martingale
mgf Moment Generating Function
PEL Posterior Expected Loss
pgf Probability Generating Function

xxv
xxvi LIST OF ABBREVIATIONS
PP Point Process
RHS Right-Hand Side
RCLL Right Continuous Left Limits
SLLN Strong Law of Large Numbers
sMG Sub-Martingale
SMG Super-Martingale
TMSO-PPRS Theory and Modeling of Stochastic Objects: Point Processes
and Random Sets
UMVUE Uniformly Minimum Variance Unbiased Estimator
WLLN Weak Law of Large Numbers
wlog Without Loss of Generality
w.p. With Probability
 List of Symbols

Q≪µ Q is absolutely continuous with respect to µ

B(X) Borel sets of the space X
DΨ[0,+∞) Càdlàg space: Ψ-valued functions defined on
[0, +∞) with RCLL
M or M(µ∗ ) Carathéodory measurable sets
n!
C nx = (n−x)!x!
Combination: the number of ways we can select x
objects out of n objects
n! P
k
C nx1 ,x2 ,...,xk = x1 !x2 !...xk !
Multinomial Coefficient, xi = n,
i=1
xi = 0, 1, . . . , n, i = 1, 2, . . . , k
CR[0,+∞) Continuous R-valued functions defined on [0, +∞)
↓ Decreasing Sequence Converging to
d
X=Y X and Y have the same distribution
X ∼ fX , F X , QX X has density fX or cdf F X or distribution QX
x! = 1 ∗ 2 ∗ . . . ∗ (x − 1) ∗ x Factorial of an integer x
F Closed subsets of R p
∀ For every (or for all)
∃ There exist(s)
⇒ Implies
⇔ If and only if
↑ Increasing Sequence Converging to
I(x ∈ A) = IA (x) Indicator function of the set A, 1 if x ∈ A, 0 if x < A
µp Lebesgue measure on R p
ω1 ∨ ω2 max{ω1 , ω2 }
ω1 ∧ ω2 min{ω1 , ω2 }

xxvii
xxviii LIST OF SYMBOLS
an
an = o(bn ) bn
→ 0 as n → ∞
1:1 One-to-one (function)
h dQ i
dµ
Radon-Nikodym derivative of Q with respect to µ
p
R, R Real numbers in 1 and p dimensions
R = R ∪ {−∞} ∪ {+∞} Extended real line
R+ , R0+ {x ∈ R : x > 0}, {x ∈ R : x ≥ 0}
Q⊥µ Q and µ are mutually singular measures
[P], [µ] With respect to measure P or µ
Z Integers, {. . . , −2, −1, 0, 1, 2, . . .}
Z = Z ∪ {−∞} ∪ {+∞} Extended integers
Z+ , Z+0 , Z+ {1, 2, . . .}, {0, 1, 2, . . .}, {0, 1, 2, . . . , +∞}
Discovering Diverse Content Through
Random Scribd Documents
interstitial tissue. Litten's65 explanation, sustained by his
experiments, is that the venous reflux, after a closure of the
pulmonary artery, is by no means necessary to the formation of an
infarction. The infarction fails if the pulmonary artery and the
bronchial artery, and those arteries lying outside the lungs, but in
circulatory connection with them—the pleural—are simultaneously
shut off. If the whole arterial supply be thus taken away, but a living
connection be maintained by means of the veins, an infarction does
not follow, while it immediately follows if, at the same time with the
open veins and closed pulmonary arteries, the collateral or
supplementary circulation be kept free. A venous reflux cannot occur
so long as a circulation in the capillaries of the lung is sustained by
collateral arterial branches. The explanation is that in an
unobstructed circulation the entire resistance which is offered to the
blood-stream in the capillaries of the lung is overcome by the
pressure existing in the pulmonary artery, which, corresponding to
the greater width of the capillaries, is much less than the pressure in
the corporeal arteries. If the pulmonary artery becomes suddenly
impermeable, the pressure in the collateral arteries, which originates
partly from the bronchial artery, and partly from those outside of,
but in connection with, the lungs, as the pleural, etc., is sufficient to
prevent a venous reflux, but not sufficient to overcome the entire
resistance in the lungs and to drive the blood beyond the capillaries
into the left auricle. Then follows an accumulation and stasis of the
blood in the capillaries and smaller veins, and hence results at first a
hyperæmia and later a diapedesis. Litten makes another important
change in Cohnheim's doctrine: he maintains that the hemorrhage
appears before the integrity of the vessel-walls is impaired.
65 Zeitschrift für klin. Med., vol. i. p. 148, Berlin, 1880.

Other fatal cases find their anatomical basis in the softening and
ulcerating processes, which while forming cavities are liable to open
vessels of greater or less size in their walls or trabeculæ.
The condition of the heart in phthisis is one which has an effect in
influencing the occurrence of hæmoptysis. The general statement by
Peacock, that the weight of the heart in phthisis, though less than in
acute diseases, is greater than that in other chronic diseases, needs
to be modified somewhat, as he did not make a distinction between
different forms of phthisis. The more the case approaches the fibroid
variety the more likelihood of some increase of size, particularly in
the right ventricle. Spatz,66 a later authority, gives as the result of his
examination that phthisis diminishes the size of the left ventricle—
that an absolutely compensatory hypertrophy of the right ventricle,
which is apparent in special cases, does not as a rule exist, although
the resultant decrease does not throughout stand in relation to the
decreased weight and volume of the whole body in phthisis. The
ratio between the depth of the left ventricle and circumference of
the aorta is diminished; and, as this is not compensated for by
hypertrophy of the walls of the ventricle, arterial tension diminishes
and the pulse becomes soft and small. The chance of rupture of
weak vessels by relatively excessive tension is thus much weakened
in the later stages of phthisis.
66 Deutsches Archiv für klinisch Med., vol. xxx. p. 154.

Another element capable of modifying the hæmoptysical features of

phthisis is claimed by Jaccoud67 as existing in the insufficiency of the
tricuspid valve, which compensates the increased tension in the field
of the pulmonary artery arising from obstruction of a considerable
part of it. The amount of blood passing from the right ventricle is
thus, by a reflux, proportioned to the area of obstruction in the
artery, and the tension is reduced so as to prevent rupture of the
weak vessels. His conclusions are based on 18 cases of
measurements of the tricuspid orifice. They varied from 111 to 130
millimeters. The evidence obtainable during life was a systolic
murmur heard at the ensiform cartilage and cervical venous reflux.
67 Clinique médicale, vol. ii. p. 346, etc.
DIAGNOSIS is mostly called for with regard to the chances of
hæmatemesis. Inspection of the blood is naturally an early point for
attention. Its bright-red color, frothy look, freedom from extraneous
matter, and its coming up by coughing are strong evidences easily
acquired of its pulmonary origin. Corroborative circumstances are
the family history of phthisis or hæmoptysis, the presence of
pulmonary, or in fewer instances of cardiac, physical signs, the
immediately premonitory symptoms spoken of before. Fever, the age
of the patient, and the continuance of the discharge of blood in its
later gradations of color and mixture of catarrhal elements,
inspection of the mouth, fauces, and larynx, would exclude those
possible sources. Each has its limitations, but together they are
conclusive as against hæmatemesis. Recent and accumulating
experience attributes some diagnostic value to the presence of bacilli
tuberculosis in the expectorated blood. Hiller68 reports 6 cases of
hæmoptysis in which the blood showed in bacilli: 3 were completely
initial. The bacilli were easily demonstrated by preparations and also
by inoculation on guinea-pigs. They have also been found in the
blood of cases of acute tuberculosis by Weichselbaum.69 Resort may
be had to the well-known tests for the presence of the elastic tissue
of the lung in sputum.
68 Centralblatt für die med. Wissenschaft, March 24, 1883.

69 Wiener med. Wochenschrift, No. 13, 1884.

As positive data for hæmatemesis we have the dark color of the

blood, its firmer clotting, greater density and want of aëration, acid
reaction, the presence of extraneous matters of food and drink, their
ejection by vomiting, and pain or uneasiness at the epigastrium. As
corroborative we have the less frequent occurrence of
hæmatemesis, the individual history of gastric disease, such as ulcer
of the stomach or presence of hepatic cirrhosis from intemperate
habits, and the history of a blow on the abdominal surface:
discharges of blood from the bowels are more likely to occur in
hæmatemesis.
Hæmoptysis may be simulated, as by scratches or cuts on some part
of the internal surface of the throat or mouth. The blood is then
likely to be thinned by secretion from the mouth. Inspection would
detect the imposture. The chance of blood from an epistaxis being
swallowed and afterward ejected by vomiting is to be remembered.
Cardiac hæmoptysis is distinguished in most cases by the presence
of symptoms and physical signs of valvular, usually mitral, disease in
a considerable degree of advancement. These are so pronounced as
to exclude phthisical disease. Other and fewer cases occur where the
hæmoptysis is the first evidence of the cardiac disease, and they
require a careful exclusion of all the features of tubercular disease,
so as to be able to arrive at a correct conclusion. There are no
conclusive considerations pertaining to the amount and character of
the blood. In the severe and copious hemorrhages there is likely to
be present a marked shock.

PROGNOSIS.—Hæmoptysis usually implies phthisis existing or

imminent, and yet it has in general a favorable effect on its course.
This applies more to its first stage than subsequently, and more to
the small than to the large hemorrhages. The gravity of the small
ones increases in proportion to their frequency. The family and
personal equation is of more importance than the mere quantity. We
may have a slight hæmoptysis and a large increase of the morbid
condition following it, and the reverse, the result depending on the
individual tolerance of and susceptibility to reaction. As in the second
stage the reactive elements are more potent, the small hemorrhages
then are less beneficial. They are the index of activity in the
destructive lesions, and yet may relieve the accompanying
congestion. The easiest appreciable symptom of the progress of the
disease is the fever. We may fail to properly interpret physical signs
because of want of familiarity with the individual case before us. If
besides more fever there be more cough, dyspnoea, and debility, the
prognosis increases in gravity. These remarks will apply with more
force to the large hemorrhages than the smaller ones, and are
guides for prognosis in all the clinical forms of hæmoptysis. In the
special clinical form, the hemorrhagic variety of phthisis, bleedings
recur often and in large quantities during years, and some of the
cases end with final recovery. The fever and constitutional irritation
give way under seemingly very unfavorable conditions. The fact that
a great part of them have no history of heredity, and that they come
on at a late period of life, may account for this, because they thus
escape the influences which heredity and age are known to impose
upon the other classes of phthisical subjects. Some interesting
conclusions have been drawn from the history of cases of profuse
hemorrhages. Pollock70 thinks that they shorten the duration of the
first stage and lengthen the duration of the second and third. Out of
his 351 cases, 204 occurred in the first three months of illness: 45
had remained in the first stage when examined, 142 having
undergone softening, while 164 had cavities. Of 286 cases of
profuse hæmoptysis classified by Williams,71 the number of cases in
the first stage was 187, and the percentage of deaths was 13.95; 65
cases were in the second stage, and the percentage of deaths was
24.61; 31 were in the third stage, and the percentage of deaths was
67.74, showing increased effect of hemorrhages upon pulmonary
structures advancing in destructive processes and upon constitutions
being progressively undermined by them. In other clinical varieties
the symptom is so clearly subordinated to the general process that it
loses its prognostic importance in the established disease. There is
an imminence of fatal hemorrhage in many of them, as in fibroid
phthisis, cancer, abscess, gangrene, and hemorrhagic infarction of
the lungs. In extra-pulmonary hæmoptysis or in that from rupture of
pulmonary aneurism there is seldom opportunity for prognosis.
70 Elements of Prognosis in Consumption, p. 139.

71 Pulmonary Consumption, p. 150.

If the condition be recognized, we can but say that the fatal attack is
liable to come at any moment. In cardiac hæmoptysis the
hemorrhage is an event coming toward the close of organic and
obstructive changes which are not much within our control. There
are minor degrees, as shown by expectoration of single small
masses of dark coagulated blood and by the absence of marked
aggravation of the symptoms, which do not prognosticate
unfavorably for the immediate, but do show impending dangers of a
future, attack. Morbid anatomy shows traces of a recovery from a
number of premonitory threatenings. The elements of a serious
prognosis are the appearance of a shock, increased dyspnoea, a
large amount of hæmoptysis, increased perturbations in the heart-
action, and increased areas of dulness or râles at certain parts of the
lung other than the usual sites of consumptive disease. These and
other evidences of constitutional initiation are not as available as in
the other varieties mentioned.

TREATMENT.—In the cases of the mildest form very little more need be
done than to keep the patient quiet. His apprehensions may require
attention. They may be allayed by assuring him that the hemorrhage
will be more of a security than a danger, because it is the expression
of a local congestion that will be relieved by the discharge. We have
found that a large dose of quinine (ten or fifteen grains) will answer
the double purpose of a nervous sedative and of controlling the
congestion and hemorrhage if the latter object be necessary. This
suggestion becomes still more applicable in the severe forms of
hæmoptysis. The dose may be repeated within twenty-four hours if
needed. If congestion be manifested by its symptoms of substernal
heat, soreness, oppression, dyspnoea, and cough to a greater
degree, and if the hemorrhage is becoming copious and the
hemorrhagic pulse developed, and the temperature elevated, the
necessity of a more active interference is evident. Absolute quiet in
bed, fresh air, a calm and equable behavior on the part of the family
or friends in attendance so that no excitement may be reflected to
the patient, are essential. The medicines selected should be such as
may control the vascular excitement, and hæmostatics. Ergot will
fulfil such indications. It has its limitations in its unpleasant taste,
but it should be pushed to the points of tolerance. Of the fluid
extract one teaspoonful should be given every hour or two until
some effect is observed in slowing the pulse or checking the
hemorrhage. If the stomach rebel, ergotin pills may be substituted in
doses of three to five grains at the same interval. Should all the
resources of ergot medication be required or the above mode of use
fail or disagree, hypodermic injections may be added. Two to three
grains of the extract of ergotin would form a proper dose, to be
repeated every one or two hours. It has been quite the exception in
our experience to have serious irritation follow the use of it in this
way. Failure in this and other uses of ergot will follow because we do
not administer it with sufficient freedom.72 Another most valuable
hæmostatic is turpentine. It should also be given freely. From ten to
thirty drops in an emulsion or in sugar may be given every two to
four hours, according to tolerance and to the threatening character
of the case. The ergot and turpentine are best alternated at intervals
of one to three hours, according to the requirements of the attack.
Some preparation of opium is often required to quiet cough—
morphine or codeine, one-fourth grain of the former and one-half
grain of the latter, repeated at intervals until their effects are
obvious. By adding the use of broken ice and the external
application of cold compresses frequently repeated, and, if time and
strength permit, the inhalation of persulphate of iron spray twenty or
thirty minims in half an ounce of water, we get a plan of treatment
adapted to the urgent cases. Some recent reports have confirmed
the confidence of the ancients in the use of ligatures. They may be
applied to both lower limbs. A dozen dry cups may be applied to the
chest. There is no occasion or time for the use of many medicines,
but if a general plan, such as the above, must be changed, acetate
of lead in doses of two grains every two hours would be an excellent
substitute, due regard being had to the possible toxic effects from
too long continuance of it in such doses; it is usual to add a little
opium to it. Gallic acid is an effectual remedy for the control of
different kinds of hemorrhages. Like ergot, it is usually given in too
small quantities. Twenty to thirty grains must be given every two to
four hours. It is better borne by the stomach, and can often be
continued longer, than the medicines above mentioned.
72 A medical friend, T. C. Minor of Cincinnati, has in his own case used three or four
drachms of the fluid at a dose, with the effect of reducing his pulse twenty beats in a
 few hours.

We have already noted ipecacuanha as one of the survivals of

ancient practice. It has had warm advocates among modern
physicians. Graves places vivisection first and ipecacuanha next in
his plan of treatment. Trousseau strongly recommended it. Peter and
the French practitioners also strongly endorse its use in the severe
forms. We have no doubt of its efficacy. It is important to exclude if
possible the existence of a pulmonary aneurism or any such source
of blooding, as there are no special means by which this can be
done. It is a good rule to use the ipecacuanha in the cases of early
or first-stage hæmoptyses. We would give it as it is given in
dysentery. Precede its administration half an hour with thirty drops
of laudanum, then give ten grains in water. If vomiting comes on,
repeat it in an hour, and again, if hemorrhage continue, in two
hours. The usual experience is that tolerance is established after two
or three doses. It has also an application in small doses of one-
quarter to one-half a grain in the milder forms, with irritative cough
and slight fever.

Graves calls attention especially to the excellent effect of opium in all

kinds of passive hemorrhage, hæmoptysis as well, but insists that it
should be given only after vivisection has been performed or when
the hæmoptysis has become rather passive, or in scorbutic and
similar cases. His direction on one occasion to a physician, in a case
of protracted bleeding of the gums, was, "Go home and give two
grains of opium immediately, and then half a grain every hour until
the bleeding stops." A combination applicable to the persistent
bleeding recurring day by day is the sulphate of magnesia made
soluble in rose-water by the free use of dilute sulphuric acid—one
teaspoonful of the former, fifteen drops of the acid, one-half to one
ounce of the rose or plain water. Many other remedies might be
mentioned, and among them atropia. After the bleeding has ceased
it is necessary to be assured as to the condition in which the lung
has been left, and to counteract, if needed, any persistence of
irritation. Fever is the most valuable evidence as to this point. If it
exist, the use of quinia and ergot had better be continued freely. A
three-grain ergotin pill about three times daily, and five grains of
quinia morning and evening, can be tolerated two or three weeks.
Local irritation should be applied if physical signs or pain warrants it.
 PULMONARY APOPLEXY.

BY WILLIAM CARSON, M.D.

DEFINITION.—Escape of blood into the pulmonary parenchyma, with

laceration of its substance.

SYNONYMS.—Hemorrhage (pulmonaire) foyer (Jaccoud); Diffuse

pulmonary apoplexy or Diffuse pneumorrhagia (Fleich); Diffuse
pulmonary apoplexy (Loomis); Pneumo-hemorrhagie (Gendrin), etc.

HISTORY.—Latour1 is quoted as being the first to use the words,

"apoplexie du poumon."2 Yet Duguet3 also quotes from Frank that
Dolocus had a long time before employed it. It is known that cases
had been described long before this, as by Corvisart in 1808, Allan
Burnes in 1809, among those of this century, and by Prosper
Martiano, Bonet, Morgagni, Haller, etc., among the more ancient
authors.4 Again, Laennec gives the weight of his authority, and
establishes Latour's use of the name, until, as the synonyms show,
modern usage has almost abandoned it. Among the multitude of
those who have treated of pulmonary apoplexy, we will have filled
the requirements of this brief historical statement by mentioning
Virchow and his pathological investigations into embolism, and also
Cohnheim,5 and later Litten's6 studies on infarction, which have
some indirect connection with pulmonary apoplexy.
 1 Histoire philosophique et médicale des Hémorrhagies, 1815, passage misquoted in
L'apoplexie pulmonaire by Duguet.

2 Op. cit., pp. 220, 222, and 224.

3 Op. cit., p. 11.

4 Duguet, p. 10, etc.

5 Untersuchungen über die Embolischen processe, 1872.

6 Zeitschrift für klinisch Medicin, Erster Band, 131.

ETIOLOGY.—Predisposing Causes.—The male sex affords

predominance of cases, because of greater liability to accidents, to
the various forms of ulcerative destruction of lung-tissue, and to
aneurisms of the aorta and pulmonary artery.

The adult age is most exposed for similar reasons. Ogston's

statistics7 support in a general way, but not with strictly technical
force, their quotation by Herz.8 Omitting the last 4 of his 20 cases (2
of which were from poisoning, 1 from fracture of skull by a fall down
stairs, and 1 from drowning), the average for males (12) was 56.4,
and 55.3 for females (4).
7 Brit. and For. Med.-Chir. Rev., vol. xxxvii., 1866, p. 459.

8 Ziemssen's Handbook, vol. v. p. 298. Ogston says (p. 465) it did not appear,
however, that any distinct rent of their substance had taken place—to any extent, at
least. "When we consider that the area of the extravasation was sufficient to involve
often one or two entire lobes, and that death was in most of them very sudden, the
cases may be adopted as showing the action of causes similar in kind, if not in
degree, to those operative in undoubted pulmonary apoplexy."

As more efficient predisposing causes than either age or sex, may be

mentioned aneurisms of the aorta and pulmonary artery, amyloid
degeneration of bronchial and pulmonary vessels, the influence of
Bright's disease in producing disease of blood-vessels, and
atheromatous diseases of the pulmonary artery.

Exciting Causes.—Penetrating and contused wounds of the chest by

their direct mechanical effect, and diseases and injuries of the brain
through the medium of the nervous system, may produce pulmonary
apoplexy, the result in the latter case being usually an infiltration or
small infarction.

SYMPTOMATOLOGY.—Pulmonary apoplexy is the least common of the

two forms of distinctive pulmonary hemorrhage, the other being
pulmonary infarction, already treated of under HÆMOPTYSIS. A
proportion of cases is associated with mitral disease in its most
advanced stages. At that time we may expect pneumorrhagia, but
whether from infarction or apoplectic laceration even the event can
only occasionally determine. In the latter, if hemorrhage makes its
appearance it will be copious and generally overwhelming; at other
times the hemorrhage may not appear, and the patient dies
suddenly with possibly other indications of the internal flow. The
physical signs cannot be relied on, for often the pulmonary tissue is
already changed by the long-continued obstruction of circulation.
Rupture of aneurism, particularly of aorta, in the great majority of
cases takes place into a bronchus, and not into the parenchyma. In
case of wounds of the contused variety a laceration of parenchyma
occurs at times sufficient to produce marked hæmoptysis. If the
blood be not ejected, there are no certain indications of what has
happened. If the case be seen immediately after the accident, such
physical signs as moist bubbling râles on the margin of an area of
feeble or suppressed vesicular murmur, possibly attended with a dull,
high-pitched percussion note over that area, would afford a strong
presumption in favor of ruptured lung and consequent hemorrhage.

The same signs later might be due to a limited traumatic

pneumonia. If the internal hemorrhage, whether traumatic or
spontaneous, has made its way through the pleura, then, if the
patient live long enough for examination, besides such symptoms as
great oppression and exhaustion, the physical signs peculiar to
pleural effusion may appear to a limited degree. This opportunity
seldom occurs, as such a rupture produces almost invariably a
fulminant and rapidly fatal result.

COURSE, DURATION, AND TERMINATION.—The course of the lighter cases is

much like that of pulmonary infarction, and that of the severe forms
too brief for observation. As to termination, it is quite possible there
are cases of laceration so limited as to allow complete recovery, but
clinical experience shows that pulmonary apoplexy is usually fatal.

PATHOLOGY AND MORBID ANATOMY.—With branches of the pulmonary

vessels weakened by long-continued heart disease, or with such
vessels as are found with chronic nephritis, a sudden increase of
tension in them from unusual effort or excitement will precipitate a
fatal rupture. Sometimes the progress of the degeneration is so
insidious and complete that a rupture may occur without obvious
exciting cause. This is also the natural history of aortic aneurism. An
examination soon after the laceration will show a mass of blood,
usually coagulated, sometimes partly fluid, lying in an irregular
cavity with walls of the lacerated lung-tissue. After a longer time the
lung-tissue beyond the walls of the hemorrhagic focus becomes
oedematous to a certain extent. A contraction of the cavity, with
change of contents, may proceed to the extent that an encapsuled
mass of very small size will remain as the final result (Rokitansky).

DIAGNOSIS from bronchial hemorrhage by the probable existence of

phthisical conditions and history. The quantity of blood ejected may
be profuse in either case, and therefore be no criterion. From
pulmonary infarction, as the other form of pulmonary hemorrhage,
by the larger amount of hæmoptysis. If there be no hæmoptysis, a
presumption would exist in favor of apoplexy in case of extreme
dyspnoea or a fatal result. The associated diseases or causes being
similar, no inference from the medical history would be reliable
except in case of injuries.
PROGNOSIS.—As we meet with it in recognizable form, the result is
almost invariably fatal. A qualification is allowed because of the
experience of such a pathologist as Rokitansky, who describes a
process of cure in a few cases. The prognosis in such would be
determined partly by the severity of the antecedent or
accompanying disease, as in heart lesions, and partly by the
increased respiratory distress, pain, exhaustion, and hemorrhage.

TREATMENT.—As has been before intimated, a case of pulmonary

apoplexy distinctive enough to be diagnosed is usually one that is
beyond the reach of treatment. Remedies that may relieve
dyspnoea, exhaustion, and hemorrhage are those to be relied on.
External irritants, as turpentine-stupe stimulants, ergot, turpentine
internally, and such other remedies as are of known effect in the
treatment of the associated heart troubles and of the incidental
pulmonary infarctions.

ABSCESS OF THE LUNG.

BY WILLIAM CARSON, M.D.

DEFINITION.—A circumscribed suppuration of the lung, resulting in a

cavity.

SYNONYMS.—Abcés du poumon; Lungenabscess.

HISTORY.—The ancients described abscess of the lung, and placed it
among the terminations of the inflammation of that organ. They
believed that if the inflammation did not resolve itself by the
fourteenth or twenty-first day its termination was to be by
suppuration. Hippocrates, Van Swieten, and others are mentioned
among those who maintained these views and consequent frequency
of such cases which prevailed until physical methods and
pathological investigation proved their incorrectness. J. P. Frank,
Bayle, and Cayol1 are given credit for a partial reversal of this
opinion. Avenbrugger,2 a pupil of Van Swieten, in describing vomicæ,
divides them into two kinds—the ichorous and the purulent. By the
purulent vomica he means an encysted abscess of the chest
resulting from the conversion of an inflamed spot into a white, thick,
glutinous, fatty matter. When these communicate with the bronchi
and discharge any of their contents by expectoration, they are called
open; otherwise, close or shut. He gives symptoms and signs
belonging to the respective varieties. Corvisart, in his comments on
these propositions, says: "In fact, the purulent vomica is always the
result of an inflammation, more or less acute, of the lung." He
makes distinctions between the various kinds of purulent vomica and
the ichorous vomica. Laennec, as in many other subjects of which he
treated, has the credit of placing this one on its modern basis, at
least so far as the frequent termination of pneumonia in abscess is
concerned. Among English writers Stokes deserves especial mention.
Abscess was the fifth and the last of the stages of pneumonia,
according to his classification. He treated largely of the perforating
abscess. Traube, Trousseau, and Leyden are among those who have
contributed largely to the elucidation of the subject. The latter has
especially claimed for this subject a more prominent place in the
literature of practical medicine, and has strongly enforced his views.
1 Nouv. Dict. de Méd. et de Chirurgie, tome xxix. p. 394; and Leyden, "Ueber
Lungenabscess," Sammlung klinische Vorträge, von Richard Volkmann, Nos. 114, 115.

2 On Percussion of the Chest, a translation of Avenbrugger's original treatise by John

Forbes, with comment by Corvisart, 1761–1808, pp. 38, 43, etc.
ETIOLOGY.—Predisposing Causes.—Everything tending to debilitate the
constitution may become a factor in the production of abscess of the
lung. Senile constitutions, Bright's disease, chronic alcoholism,
diabetes mellitus, and insanity are some of the predisposing causes.

Exciting Causes.—These may be divided, as in the case of gangrene,

into the pulmonary, or those originating in the lung or pleura, and
the intra-pulmonary, or those originating outside of the lung or
pleura. Among the former are included pneumonia and empyema,
perforating and discharging into the lung, or one variety of Stokes's
perforating abscess, pulmonary apoplexy, and suppurating bronchial
glands, opening up a passage through the lung and bronchial tube.
Either croupous or catarrhal pneumonia may be associated with or
terminate in pneumonia. Among the latter, or extra-pulmonary class
are included cases of embolism from the right heart, producing
infarction, or from the systemic veins. These emboli carry with them
the productive capacity of suppuration. Abscess external to thoracic
walls, as in deep-seated mastitis, will at times perforate the walls
and enter the lung. Abscesses of the liver not infrequently perforate
the diaphragm, and are discharged through the lung.

Foreign bodies in the bronchi may ulcerate through them and

produce suppuration of the lung, which may finally open a way
externally through the chest-walls.

SYMPTOMATOLOGY.—The symptoms of abscess of the lung, as may be

inferred from the enumeration of causes, are divisible into two
categories—one including those symptoms with which the abscess
may be associated, but which do not necessarily prognosticate it;
and the other including the symptoms which indicate the abscess as
a fact accomplished. In the simplest and most frequent clinical form,
that following pneumonia, the early symptoms would be those of a
severe and irregular form, as shown by very troublesome and
uncontrollable cough or unusual pain or respiratory embarrassment,
high fever, but at that time fairly typical pneumonic temperature,
great prostration, etc. These may all diminish in due time, and
mostly do without suppuration following. A return of pain, dyspnoea,
fever, and general distress should awaken suspicion, yet they may be
the result of an extension of pneumonia to other portions of the
lungs. Rigors and sweats and increased depression would point to a
suppurative process and under such circumstances to the lung as
the locality. We cannot, then, positively predict an abscess. It is
suspected when a more or less copious eruption of purulent
discharge occurs suddenly, and sometimes the discharge is so
abundant and pus-like that any other alternative than abscess is very
remote; at other times the discharge is small in quantity. The proof
of physical cavernous signs is the final step. This is often difficult. A
slightly greater increase of lung-density, probably at the middle or
upper part, with imperfect bronchial breathing, the appearance of a
few moist râles or crepitus, the gradual increase of these and
merging into coarser crepitus, and revelation of more or less of the
cavity signs, is the physical history of many cases of abscess of the
lung. Others have a much more pronounced course, such as the
cases of so-called gangrenous abscess—a sort of connecting link
between gangrene and abscess of the lung. The breaking down of
tissue is ostensibly very sudden, and the cavernous signs are very
soon unmistakable. Other clinical forms are the pyæmia, to be
distinguished by the antecedent history, which will reveal a source
for infectious emboli. The abscesses are usually multiple and small,
so that their precise locality cannot be made out. The proof of
infectious transportation is sudden pneumonic symptoms, as pain,
tinged and finally purulent expectoration. Rupture into pleura may
occur and produce empyema. Rupture of hepatic abscess and
discharge through the lungs is also a clinical form shown by this
antecedent event, pointing to hepatic inflammation. The egress of
the pus is sometimes through a narrow track, and not by a reservoir
within the pulmonary tissue; at other times the lung is really
excavated. The discharge of pus is usually copious and paroxysmal.
Leyden recognizes as his third class a form of chronic abscess, or
one coming on during a case of chronic pneumonia and bearing
great resemblance to a variety of phthisis. Its general symptoms are
much the same as in the acute variety, differentiated by the element
of time.

COURSE.—It may be said, in a general way, that the etiology has

much to do with its course. If the cause be pneumonia, the course
will be such as the detail of symptoms already given shows. In some
unknown way the natural course of the disease is interrupted, and
what promises to be an average case is followed by the
characteristic features of abscess. If pyæmia be the precedent
condition, a peculiar form of pneumonia, embolic in origin, appears,
and abscesses again follow. Greater septicity and rapidity of
destruction are probable sequences. The perforating abscesses are
subject to modifying influences of mechanical effect, such as
gravitation and the resistance of tissues, and have their peculiar
course, which is often marked by great chronicity.

TERMINATION.—In the course of seventeen years the reports of the

Cincinnati Hospital show that there have been 6 cases of abscess of
the lung treated there. Of these 4 died and 2 were discharged as
improved. These figures show the infrequency of such cases, and
also represent a greater mortality than probably occurs in the non-
hospital class. We know of no large statistics which show what is the
percentage of recoveries. Our own experience in private practice
gives a majority of recoveries. They were cases following typhoid
fever, croupous and catarrhal pneumonia, and hepatic abscess. A
termination in a chronic cavity now and then happens: perforation of
the pleural cavity, with subsequent pyo-pneumothorax, discharge
externally through an intercostal space, or even extension into the
abdominal cavity, are among the actual events of such abscesses.

DURATION.—The duration of an ordinary case is subject to wide

variations between one and six months. A few cases are recorded of
several years' duration. Previous constitutional condition has much to
do with this element. The degree of infectiousness in the pyæmic
class is important as to time. The abscesses become a subordinate
condition in the fate of the case. In this connection we may also
refer to Leyden's third variety, a so-called chronic abscess.

PATHOLOGY.—A close parallelism, etiologically and otherwise, is

observable up to certain points between gangrene and abscess of
the lung. Both are products of, or associated with, pneumonia. That
which finally determines whether the result shall be gangrene or
abscess is unknown to us. In the article on GANGRENE OF THE LUNG
some investigations are referred to which point to a probable
solution in the existence of specific forms having special
pathogenetic force. The tendency of experimental and clinical
investigations is to connect the suppurative process closely with the
product of specific germs. Ogston in 65 cases of acute abscess found
micrococci present in all of them. Obstruction of blood-vessels in the
centre of the pneumonic area or on the margin of the abscess walls
is an important anatomical element in the production of abscess,
and it is claimed that it is often due to colonies of micrococci within
their calibre; so that it is probable that there are both mechanical
and biological or chemical influences at work. If the view of the
zymotic and infectious character of pneumonia be tenable, the
contingency of an abscess developing in its course would seem not
very remote. Yet the proportion of cases of abscess from pneumonia
is not more than 2 per cent. Leyden's high authority supports the
idea of the essential and specific differences in the chemical and
morphological peculiarities of gangrene and abscess of the lung, but
the subject is as yet on a hypothetical basis.

MORBID ANATOMY.—The fresh cavity, generally in the upper lobe, has

rough, ragged, and irregular walls, and may have bridles of the more
resistant structures, as bronchi and vessels, crossing it. Such a cavity
is quite likely to contain portions of undissolved parenchyma or more
or less malodorous pus. The older cavity becomes smoother walled,
and of more regular limits and cleaner contents. A gradation from
granular hepatization through congested to crepitant tissue is almost
uniform in the varieties of abscess, whether simple or pyæmic. In
addition, some peculiarities are observable in the latter. These are
usually several, varying in size from a pea or less to a walnut, some
round and others wedge-shaped; others lying superficially and
forming slight elevations on the pleural surface. In proportion to the
curative progress the cavity will contract and disappear, occasionally
leaving behind a cicatricial mark. A lining pus-secreting membrane
will sometimes form, resulting in such a limitation of morbid action
and such a disappearance of reactionary symptoms as to make the
disease entirely local, but quite chronic.

DIAGNOSIS.—The more or less sudden and copious expectoration of

pus, without a specially offensive odor, in the course of a case
having up to that time the history of a pneumonia, would be
considered as due to the development of an abscess in the lung.
Some degree of fetor in breath and expectoration is observed, but it
is far different from that of gangrene. The detection of the débris of
lung-structure in coarse particles, and the microscopic discovery of
elastic lung-tissue, are important diagnostic points in
contradistinction from the solution of tissue that gangrene usually
effects upon the parenchyma of the lung. According to Leyden's3
very complete investigations, the microscope reveals fatty crystals,
mostly in roundish fragments, of the size of the epithelium of the
lung and of a brilliant structure; pigment-débris of a yellowish-brown
or brownish-red color; hæmatoidin and bilirubin, which Traube
thought were due to hemorrhagic infarction, but which Leyden has
observed in all of his cases; and, lastly, micrococci, in the well-
known form of the round, granular micrococci colonies, which differ
from those in gangrenous fragments in that they show very little
movement and do not give the iodine reaction.
3 "Ueber Lungenabscess," Volkmann's klin. Vorträge, p. 994.

Difficulties of diagnosis arise in the case of an empyema discharging

through the bronchi, or of an encysted empyema discharging
through the third or fourth intercostal space in front; also, between
abscess of the superior portion of the liver and one in the base of
the lung, or between the latter and a pyo-pneumothorax. Very
careful study of the history in each case is of the first importance.
Where this is not attainable the difficulty is often much increased. In
the case of the empyema the discharge is more profuse at each
time, the whole amount in a given period is much greater, and the
time of opening is much delayed beyond that of the pneumonic
abscess. Trousseau gives the case of a child who brought up for
more than six months 200 grammes of pus daily. He makes children
an exception to the rule as to the late opening of the pleural
abscesses. In the encysted empyema discharging either internally or
externally the difficulties are greater. A portion of the lung-tissue
may be so near behind the deposit of matter as to make the physical
signs confusing if the pus has opened externally. Some of these and
of the interlobular deposits it is almost impossible to diagnose.

In hepatic abscess opening into the lung and bronchi the discharge
is copious, dirty brown, paroxysmal, and will generally, on careful
observation, show the bile color or its chemical reactions or some
microscopic débris of the liver. In Leyden's third class, or the chronic
abscess arising in the course of chronic pneumonia, the history is so
much like that of some forms of phthisis as scarcely to serve in
diagnosis. He thinks there are some macroscopic and microscopic
appearances which may serve for diagnosis. There are in the
expectoration dark and compact pieces of greenish-black color, not
unlike plugs of pus, and larger, black-pigmented fragments of
parenchyma, from a pin's head to a hempseed in size.
Microscopically, they consist of a close and strongly-pigmented
parenchyma, which seldom reveals alveolar structure. They show
fatty degeneration and cholesterin plates. This class of cases is
mostly without fever. The application of the bacilli-tuberculosis test
would seem to offer some assistance in diagnosis.

PROGNOSIS.—A grave prognosis may be formulated if there be a

history of feeble constitution, and especially if it be further impaired
by habits of intemperance, if the patient belong to either extreme of
age, if there has been a recent debauch, or if there be wide variation
from the typical form of pneumonia. Variations will be shown in such
a complexus of symptoms as follows: fever of low grade, subject to
extremes in range; feeble and frequent pulse, but not so marked as
in gangrene of the lung; dyspnoea, objective and subjective; typhoid
depression; tongue dryish; delirium; copious and fetid or difficult
expectoration; physical signs of extensive lesion, such as a large
cavity with a large outlying pneumonic area. A favorable prognosis
would be conditioned on the appearance of a fewer number of these
symptoms or on their evolution in a milder form.

The capacity of the patient to endure a long-continued suppurative

discharge is principally determined by his natural vigor and his ability
to assimilate food, other elements, such as extent of injury to the
lung, being the same. A well-defined superficial cavity would be
more favorable, because within surgical relief.

In the pyæmic variety the force of the infectious element will

determine largely the result. Chills and sweats are important
prognostic elements in such a case.

In the secondary abscesses of either the empyematous or hepatic

variety prognosis is grave—more so in the latter than in the former,
because surgical procedure would be more promising in the former,
and because of the implication of an organ so liable to destructive
inflammation as the liver. A long and tedious course of suppuration is
possible in either. The dangers in an established abscess arise from
liabilities to septic infection and exhaustion consequent on want of
reparative power and persistent suppuration.

TREATMENT.—The treatment of abscess differs little if at all from that

of gangrene of the lung. The tendencies of the two diseases toward
exhaustion and infection are similar, but are less pronounced in the
former. The same remedies are necessary in both, such as
stimulants, tonics, antiseptics, anodynes, and expectorants
internally, inhalations and drainage externally; brandy and malt
liquors as stimulants; nourishing and concentrated food at frequent
intervals; quinine as tonic and antiseptic; carbolic acid and
turpentine as most valuable antiseptics (the latter being also an
excellent stimulant); eucalyptus in cases of profuse as well as fetid
discharge; carbonate of ammonia, senega, as expectorants;
morphine and codeine or anodynes to control cough; carbolic acid
for inhalation; and in cases of definitely localized cavities a free
opening to be made with antiseptic injections.

Successful cases of surgical interference are reported, and such

treatment is now recognized as proper when the system is giving
way under septic poison, evinced in chills, sweats, and great
prostration, where the purulent discharge is fitful and imperfect, and
where the physical signs are clear enough to show the locality of the
abscess.

GANGRENE OF THE LUNG.

BY WILLIAM CARSON, M.D.

DEFINITION.—Putrid necrosis of the lung-tissue.

SYNONYMS.—Lungenbrand, Gangrene du poumon, Gangræna

pulmonum.

HISTORY.—By common consent, Laennec has the credit of first

identifying, naming, and classifying gangrene of the lung as a
distinct disease; yet Lieutaud1 in 1707 describes imperfectly a case
of gangrene of the lung in a child: "the right lung, within and
without, appeared entirely putrid." Bayle2 is considered, in his
section on his fourth variety of phthisis (phthisis ulcereuse), to have
described a rather chronic form of gangrene of the lung. Morgagni,
Boerhaave, Stoll, J. Frank, and Cullen considered gangrene as one of
the terminations of peripneumonia.3 Laennec's development of the
subject has only in a few directions been enlarged. His classification
is universally adopted. His description is adopted generally as the
most complete. There have been, however, controversies on
different points, such as the relation of pneumonia and of the
obstruction of the vessels to gangrene of the lung.
1 Historia Anatomica Medica, 1787, Obs. 329, cited by Louisa Atkins, 1872.

2 Bayle, G. L., Recherches sur la Phthisis pulmonaire, 1809–10, p. 30.

3 I. Straus, Nouv. Dict. de Méd. et de Chir., p. 403, etc.

In the pathology and etiology of gangrene Virchow's investigations

on embolism and thrombosis opened up important relations; in
diagnosis, Traube and Leyden and Jaffee; in medical treatment, also
Traube; and in surgical treatment, Haley and Lawson (1879),4 S. C.
Smith (1880), E. Bull (1881), Fengar and Hollister (1881), Mosler
and Voght (1882). The antecedent development of pulmonary
surgery, through important work done by Mosler, Pepper, and others,
had prepared the way for special applications of it to gangrene and
abscess of the lung. Spencer Wells claims to have suggested similar
proceedings nearly forty years ago.
4 Lungen Chirurgie, Mosler, xx. p. 67.

ETIOLOGY.—Predisposing Causes.—Constitutional weakness is a

common predisposing influence: it may be a primary condition, but
is more often secondary or dependent on some recently-acting
debilitating cause, as typhoid fever, chronic lung disease, diabetes,
etc. Chronic alcoholism is a cause which, besides its effect on the
system at large, may add a special one on the lungs in producing
hyperæmia or drunkard's pneumonia.
Of 46 cases we have collected mostly from the Vienna Hospital
report, the youngest was nineteen years old and the oldest was
forty-seven years. Lebert5 has collected altogether 60 cases, 32 of
his own and 28 of others: 19 occurred between twenty and thirty
years, and 1 between thirty and forty. Huntington6 gives 32 cases
from the Massachusetts General Hospital Record between 1857 and
1875: 9 were between twenty and thirty years, and 12 between
thirty and forty; the youngest was ten years old and the oldest sixty-
four. It is noticeable that these figures coincide largely with those
showing the incidence of phthisis. Louisa Atkins7 gives, as the
youngest ages among all the varieties, one of three months and
another of two months.
5 Klinik der Brustkrankheiten, vol. i. p. 827.

6 Boston Med. and Surg. Journal, vol. xcv. p. 486.

7 Gangræna Pulmonum bei Kindern, 1872.

Of the 46 Vienna Hospital cases, 43 were male and 3 female.

Huntington's cases were males 24, females 8. Of Lebert's own 32
cases, 22 were males; of the 32 others summarized by him, in 4 sex
was not mentioned, and of the remainder 17 were males and 11
females. These figures show the large predominance of males in the
liability to attack.

Exciting Causes.—They may be classified as pulmonary and extra-

pulmonary. The influence of the alcoholic habit has been referred to
above among predisposing causes: debauches are a frequent
antecedent, especially in hospital cases, by means of resulting
pulmonary hyperæmia and drunkard's pneumonia. Its association
with croupous pneumonia may be assumed as settled after some
warm disputes. The pneumonia of Bright's disease and putrid
bronchitis are occasionally causative; bronchiectasies result in it not
unfrequently. Extension of diphtheritic inflammation from the
tracheal and bronchial mucous membrane is another form. The
catarrhal pneumonia secondary to measles may produce it in
children.

Embolism is the most frequent cause in the class of extra-pulmonary

causes. It may be mechanical or infecting. A bronchial artery may be
plugged so as to produce a gangrenous slough from mechanical
cutting off of nutrition. Embolism of the pulmonary artery branches
is more frequent, and by bringing about infarction and apoplexy may
produce gangrene. Of the infecting variety may be mentioned emboli
from the peripheral veins, as in surgical or uterine phlebitis, or from
cerebral sinuses secondarily involved from otitis. Other causes acting
from without on the lungs are foreign bodies, as particles of food
passing beyond the trachea into the lungs, as in case of the insane
or drunkards, and blows on the walls of the chest. These latter are
capable of producing not only the ordinary phenomena of contusion-
pneumonia8 but gangrene, and without evidence of external injury
or fracture of the ribs.
8 M. Litten, p. 26, vol. v., Zeitschrift für klinische Medicin.

SYMPTOMATOLOGY.—Gangrene of the lung is the termination of a

process the beginning and progress of which are not declared or
cannot be followed through characteristic symptoms. Even its final
occurrence may remain unknown if a communication be not
established with a bronchus, which event is followed by the true
symptoms, the expectoration and its odor. Whatever symptoms
occur previous to that event may occur independent of it. Adopting
Lebert's dictum,9 gangrene of the lung is not a pathological unit. As
its pathogenesis varies, so does its symptomatology. A feature
common to its several varieties is marked constitutional depression
and variations from the typical form of the disease in which it occurs.
If pneumonia, croupous or catarrhal, be the precedent or associated
disease, it will be marked by soft and feeble and frequent pulse,
restlessness, dulness or distress of countenance, more or less
cyanosis, cool and relaxed skin, possibly delirium, dry tongue,
unusual dyspnoea and pleuritic pain, copious prune-juice
expectoration, irregular or non-typical temperatures. Along with
these functional variations occur some in physical signs, as a lesser
amount of dulness or of bronchial breathing, indicative of less
structural density and corresponding exudation. A case with such an
evolution may afford a presumption of an outcome in gangrene, but
appearance of the characteristic expectoration and fetor is necessary
to exclude it from irregular forms of pneumonia, which have no such
termination. The same general remark applies to the cases of
gangrene in bronchiectasic cavities. Perhaps some aggravation of the
general condition may excite apprehension, but the characteristic
phenomena of expectoration, odor, etc. must decide. If the cause be
of embolic origin, we may expect some suddenness and perhaps
shock in the beginning, and later the evidences of a more limited
inflammation of the lung-tissue, such as circumscribed dulness and
modified respiratory sound, which finally end in those indicating
destruction of lung-tissue.
9 Op. cit., p. 803.

The macroscopic characters of the expectoration are those of a

putrid or fetid liquid of varying shades of color, ashy gray, dirty
green, or greenish-yellow, prune-juice, or more or less hemorrhagic.
The odor, which is at first so fetid and penetrating, often disappears
after the expectoration has been standing a while. It is separable, as
first described by Traube,10 into three principal layers. The
uppermost, covered with a layer of foam, consists of, first, dirty
green, crumbling, confluent lumps; second, of larger, homogeneous,
green muco-purulent masses; and, third, of whitish-gray,
transparent, mucous masses. The second layer is formed of a
colorless fluid. At the bottom is a fine yellowish-white sediment.
Microscopically are found fat-acid crystals, many large fat-drops, and
finely granular débris, masses of free, black pigment. It is said
elastic tissue is nowhere to be found, but to this statement there are
no doubt exceptions. Other bodies have been found by Leyden and
Jaffee,11 which they named Leptothrix pulmonalis. Kannenberg,12
besides the above forms, found constantly infusoria of the family of
monads in the sputa of 11 out of 14 cases of pulmonary gangrene.
He considers them peculiar to the processes of decomposition in the
lungs.
10 Gesammte Beiträge zur Pathologie und Physiologie, Zweiter Band, p. 452, etc.

11 Deutsches Archiv für klin. Med., Band ii. pp. 488, etc., "Ueber putride Sputa."

12 Virchow's Archiv, Band lxxv.; Zeitschrift für klinische Med., Band i. p. 228.

PROGNOSIS in general is unfavorable. Individually, the gravity of the

case is determined by the evidence bearing on previous habits and
constitution; by the violence of the onset, as shown in prostration,
severe pain in the chest, dyspnoea, persistent and violent cough,
delirium, feebleness and softness of the pulse; by the variations
from the typical standards of croupous, or especially catarrhal,
pneumonia, such as greater amount of septic or infectious or
typhoidal element, non-typical and low temperatures in the early
stages and also in the stage of disintegration; by the amount of the
latter as shown in the physical signs of extensive lesion and in the
amount of gangrenous sputum; and by the irritant effect of this in
producing bronchial catarrh, and consequent catarrhal secretion,
which may of itself become an element of danger in a system
already much prostrated. Favorable prognosis is allowable when
these conditions are being gradually reversed.

PATHOLOGY.—The pathology of gangrene of the lung is scarcely more

settled than it was forty years ago, when Stokes13 published his
eighteen propositions, embodying his experience. Obstruction of
vessels and inflammatory exudations are present as important
pathological conditions, but of themselves are not pathogenetic of
gangrene. Other, as yet unknown, elements of putrefactive agency
are present. Leyden and Jaffee's observations and those of
Kannenberg have been mentioned14 as efforts to throw light upon
the pathogenesis of gangrene of lung, but how far the bodies
described by them precede, coincide, or follow the familiar clinical
phenomena are undetermined questions. Filehne,15 in his
experiments to determine the reason of the almost universal
absence of elastic fibres in the expectoration of patients with
gangrene of lung, comes to the conclusion that there is a ferment
which, acting under alkaline conditions, destroys the fibres. The
agents of this ferment he does not try to determine. Stokes
anticipates the tendency of modern experimental pathology by
announcing as an alternative proposition that "a process of
putrefactive secretion precedes in many cases the death of lung."
The constitutional debility which is so early a symptom prepares the
way for such an invasion. In reference to the relation between the
septic material and thrombosis in gangrene, Kohler16 affirms that the
septic material produces the fibrin-ferment, and thereby capillary
thrombosis. Recklinghausen thinks that a special material capable of
exciting coagulation has not yet been found in gangrenous
substances, but that there may be several factors, such as anæmia,
changes in the vessel-wall, imbibition with foreign substances, etc.
Other experiments17 and views point toward the conclusion that
there are substances formed in various diseased conditions which
have the power of ferments and of producing coagulation of blood in
the lesser circulation.
13 Dublin Quarterly Journal Med. Science, Feb. 1, 1850.

14 Op. cit.

15 "Sitzungsbericht der Phys. Med. Soc." in Erlangen Schmidt's Jahrbucher, 1877, No.
7.

16 Recklinghausen's Handbuch der Pathologie, p. 136.

17 Wooldridge, Du Bois-Reymond, Archiv Centralblatt für med. Wissenschaften, No.

41, vol. xi. 1874, p. 734.

MORBID ANATOMY.—The circumscribed variety, as it usually appears, is

a cavity irregular in outline, with ragged walls, made so by the
unequal invasion of the lung-tissue by the gangrenous process.
Sometimes the cavity itself contains loose fragments of lung-tissue,
or the contents may be of a dirty, greenish, or brownish color, with
some of the odor of the expectorated matters. If the case has been
a chronic one, the walls are smoother, with a more or less formed
lining membrane and the contents of a less characteristic color or
odor. The cavity is usually in the right lung, and in the middle or
lower portion. There is much variety of statement on this point. The
tissue immediately adjacent to the gangrenous walls shows more or
less of the products of catarrhal pneumonia. The vessels terminating
in the walls are obstructed by coagula. If the gangrenous part come
to the pleural surface, belonging to what Fournet calls the superficial
variety, it may produce adhesions there, or it may result in
perforation, with the result that we have the products of pleuritic
inflammation united with the contents of the gangrenous cavity. In
multiple foci, some will show the less advanced stages of disease,
such as incomplete softening and local inflammatory lesions. In the
diffused variety the lesions are sometimes described as being the
same except in extent. The demarcations, however, from
surrounding tissue are not as well marked; the tissue is soft,
breaking easily under pressure, sometimes oedematous, dark or
dirty black on surface or on section of lung; at other times the
surface is mottled with lighter-colored patches. A secondary result is
the production of bronchitis by the contact of the irritant
expectoration from the gangrenous cavity.

DIAGNOSIS.—Abscess of the lung is the disease most likely to be

confounded with gangrene of the lung. In the former you do not
have the same amount of profound constitutional depression; the
symptoms have a more frank expression, as it were, because of
better precedent conditions usually; the first eruption of matter from
the abscess is more copious and sudden than the usual manner of
expectoration of gangrenous sputum; the appearance of the
contents of the abscess is that more nearly of healthy pus, though
the latter has at times a dark dirty brown or hemorrhagic look; the
separation into layers is not apt to occur; the odor is usually not so
fetid; microscopically, elastic fibres are much more abundant in
abscess than in gangrene of the lung. The cavernous physical signs
are not reliable in either disease. There are cases in which it is
impossible, and in which it is of no practical importance, to make a
diagnosis between gangrenous abscess and pulmonary gangrene.
Gangrene supervening in phthisical cavities is distinguished by the
history of a chronic pulmonary disease in which a cavity has been
previously marked. It is phthisis advanced to the destruction of
tissue plus the unknown gangrenous element which has found a
lodgment in a favorable place. Stagnation of cavity contents,
depression of system, etc. are favoring conditions. The same
remarks apply to bronchiectasic cavities and putrid bronchitis,
physical signs in the latter being additional points of difference.

Our experience proves that the essential SYMPTOMATOLOGY of gangrene

of lung, except the débris of lung-structure in the sputum, can occur
where there was no gangrenous destruction, either circumscribed or
diffuse, no bronchiectasic cavities or bronchial dilatations, and no
phthisical cavity. In the case alluded to the gangrenous odor and
general characters of the sputum and the separation into layers and
the consolidation of tissue were present, but the post-mortem
afforded no explanation of the fetid expectoration. It was a case of
debauch and alcoholic excesses and exposure.

COURSE, DURATION, AND TERMINATION.—The course of this disease is

essentially an acute one. Whatever the early condition be, the
gangrenous element hastens its progress, as in cases beginning with
the phenomena of acute pleuro-pneumonia. A pre-existing phthisical
cavity will take on acute phases, also the septic element will be
reinforced, and, as indicated in the enumeration of symptoms,
clinical irregularities will be introduced. The early prostration is
increased, and the patient dies from exhaustion after a period
varying from three days to six weeks from the time the gangrene
became manifest. Various complications, such as pleuritis and
perforation of pleura and pneumothorax, hemorrhage, or extensive,
diffused gangrene may cut short the patient's life within the average
period. Occasionally the gangrenous cavity becomes chronic and the
patient may live for months in imperfect health. The termination of
the circumscribed variety is usually death. Diffuse gangrene is
invariably fatal.

TREATMENT should be directed, first, to the known precedent states of

constitutional weakness, bad habits, etc., which lay a foundation for
unhealthy inflammatory processes, and for the pathogenetic
elements that bring about the gangrenous and septic and exhausting
conditions; second, to the special symptoms, such as severe pain in
the side, harassing cough, dyspnoea, etc. Remedies of the first class
are quinia, turpentine, early alcoholic stimulation, carbonate of
ammonium, antiseptics, as carbolic acid, etc. An anticipation from
the beginning of any irregular form of acute pulmonary disease of its
termination in gangrene is impossible, and hence early treatment is
necessarily general. It would be properly confined to the use of
quinia, which would be useful, either before or after the gangrenous
element had developed, in small doses frequently given, unless high
temperature or the septic process indicate the use of large ones.
Turpentine internally is desirable in proportion to the infectious or
typhoidal character of the attack. Its use by inhalation is beneficial
when gangrenous destruction has already taken place. Brandy or
whiskey in moderate and frequent quantities, one-half to one
tablespoonful every three hours, and carbonate of ammonium, meet
the requirements which the tendencies to debility indicate. Milk,
milk-punch, beef and chicken extracts should be given in the
intervals between the administration of medicines. This general plan
is applicable throughout the pre- and post-gangrenous stages. In the
symptomatic treatment pain and a general respiratory distress often
demand attention. Opium is then useful, both in relieving pain and
moderating dyspnoea and cough. As these symptoms are often
urgent in the later stages of gangrene of the lung, the treatment of
the disease harmonizes both in its constitutional and symptomatic
aspects. Prescriptions can also be readily prepared which contain
remedies that have a decided effect in correcting the fetor of the
breath and expectoration, and thus evincing an influence on the
putrefactive process or ferment, which has become the prominent
feature of the disease. The author recommends that carbolic acid, in
the dose of one grain every four hours, be given for that purpose,
and also its use by inhalation. Assistance can sometimes be given by
putting the patient in such a position on the side as to promote the
emptying of the cavity.

Tapping a gangrenous cavity and the introduction of drainage-tubes

may be successfully resorted to. The cases suitable for such surgical
treatment have been described by Fenger and Hollister18 as those
where, "the presence of a gangrenous or ichorous cavity having
been ascertained, it is found that notwithstanding an outlet through
the bronchi for a portion of the contents of the cavity, it steadily fills
up again; the partial evacuation does not relieve the patient; the
infection of the healthy portions of the lung from the decomposed
contents of the cavity has commenced or is evidently about to take
place; the breath and expectoration continue fetid; absence of
appetite; increasing weakness, with or without fever, etc." For the
steps of the operative procedure reference may be made to the
complete directions given by the same authors or to works on
surgery. The double opening advised by the above authors would be
the most efficient plan.
18 Amer. Journ. Med. Sci., Oct., 1881, p. 385.

CROUPOUS PNEUMONIA.

BY ALFRED L. LOOMIS, M.D.

SYNONYMS.—English and American: Acute sthenic exudative
pneumonia; Primary lobar pneumonia; Vera peripneumonia; and
Pneumonitis. Fr. Fluxion du poitine and Fièvre pneumonique; Ger.
Pneumonia lobaris and Lobäre Lungenentzündung.

DEFINITION.—Croupous pneumonia is an acute general disease with a

characteristic local pulmonary lesion. Anatomically considered, it is
an acute inflammation of the vesicular structure of the lungs,
resulting in infiltration of the alveoli, with inflammatory products,
which renders them impervious to air. This condition is known as
hepatization.

HISTORY.—Until the time of Laennec, pneumonia and pleurisy were

described as one disease. Hippocrates said that pleurisy was "a
disease quickly fatal, and characterized by sputa of various colors."
Although these two diseases were undifferentiated, accurate
descriptions of the lesions and objective signs of pneumonia have
come to us from the earliest medical writers.1
1 Thucydides, The Plague at Athens, B.C. 430.

Much of the early history of this disease is interwoven with the

detailed accounts of a great variety of pulmonary symptoms which
occurred in the epidemics and plagues which prevailed in Eastern
Europe in the centuries just preceding the Christian era, and in
Western and Southern Europe during the sixteenth and seventeenth
centuries. The black death has been regarded by some as an
epidemic of pneumonia. While it is probable that in most of these
epidemics the lung was early involved, and that its implication
hastened death, yet no proof exists to sustain the belief that they
were other than epidemics of typhus fever, dysentery, and those (as
yet unknown) fevers which collectively were named plagues. That
many of these plagues were complicated by pneumonia there is
scarcely room for doubt.
French investigators were the first to separate the pneumonic
process from all other morbid processes which occur in the thoracic
organs. Valsalva, Morgagni, and Boerhaave gave accurate
descriptions of pneumonia, but they did not sharply distinguish it
from pleurisy. Bichat and Pinel separated collapse of the lung
attending pleuritic effusion from inflammation of the lung-
substance.2 Laennec was the first to draw the line sharply between
pneumonia and pleurisy, and to him, more than to any other
observer, is due the credit of describing pneumonia as a distinct
disease. With his labors begins a new era in the history of pulmonary
inflammations.
2 Nos. phil., ii., Pinel.

Grisolle's work3 is especially valuable in statistics relating to the

climatic element in the development of pneumonia and its
comparative frequency among different races. The elaborate treatise
on the geographical distribution of pneumonia by Ziemssen has
furnished data for a more accurate knowledge of its geographical
boundaries. Following in the footsteps of Laennec, Chomel, Stokes,
Addison, and many laborers of our own day have furnished the
material from which the clinical and pathological history of
pneumonia is now being constructed.
3 Traité de la Pneumonie.

Recently, Jürgensen has strongly advocated the infection theory of

pneumonia, and has presented strong arguments in support of his
opinions. Sturges of London and Cohnheim4 advocate the use of the
term pneumonic fever, and the former gives a most interesting
general comparison between it and the affections which he regards
analogous to it. Careful pathological experiments have recently been
made by Heidenhain, Sommerbrodt, Schuppel, and Klebs5 with a
view to establish the germ-theory of pneumonia.
4 Leçons de Chir. méd., 1877, p. 17.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

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

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com