Download the full version of the textbook now at textbookfull.

com

R Data Mining Implement data mining

techniques through practical use cases and
real world datasets 1st Edition Andrea
Cirillo
https://textbookfull.com/product/r-data-mining-
implement-data-mining-techniques-through-
practical-use-cases-and-real-world-datasets-1st-
edition-andrea-cirillo/

Explore and download more textbook at https://textbookfull.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Data Mining and Data Warehousing: Principles and Practical

Techniques 1st Edition Parteek Bhatia

https://textbookfull.com/product/data-mining-and-data-warehousing-
principles-and-practical-techniques-1st-edition-parteek-bhatia/

textbookfull.com

Data Mining Practical Machine Learning Tools and

Techniques Fourth Edition Ian H. Witten

https://textbookfull.com/product/data-mining-practical-machine-
learning-tools-and-techniques-fourth-edition-ian-h-witten/

textbookfull.com

Data Mining and Big Data Ying Tan

https://textbookfull.com/product/data-mining-and-big-data-ying-tan/

textbookfull.com

Gender and Islam in Indonesian Cinema 1st Edition Alicia

Izharuddin (Auth.)

https://textbookfull.com/product/gender-and-islam-in-indonesian-
cinema-1st-edition-alicia-izharuddin-auth/

textbookfull.com
Environmental Human Rights and Climate Change Current
Status and Future Prospects Bridget Lewis

https://textbookfull.com/product/environmental-human-rights-and-
climate-change-current-status-and-future-prospects-bridget-lewis/

textbookfull.com

Pediatric Retinal Vascular Diseases From Angiography to

Vitrectomy Ulrich Spandau

https://textbookfull.com/product/pediatric-retinal-vascular-diseases-
from-angiography-to-vitrectomy-ulrich-spandau/

textbookfull.com

Oxford Handbook for Medical School 1st Edition Kapil

Sugand

https://textbookfull.com/product/oxford-handbook-for-medical-
school-1st-edition-kapil-sugand/

textbookfull.com

Learning Spring Boot 2.0 - Second Edition: Simplify the

development of lightning fast applications based on
microservices and reactive programming Greg L. Turnquist
https://textbookfull.com/product/learning-spring-boot-2-0-second-
edition-simplify-the-development-of-lightning-fast-applications-based-
on-microservices-and-reactive-programming-greg-l-turnquist/
textbookfull.com

Receptor Receptor Interactions in the Central Nervous

System Kjell Fuxe

https://textbookfull.com/product/receptor-receptor-interactions-in-
the-central-nervous-system-kjell-fuxe/

textbookfull.com
Biotechnology Fundamentals, 3rd Third Edition Firdos Alam
Khan

https://textbookfull.com/product/biotechnology-fundamentals-3rd-third-
edition-firdos-alam-khan/

textbookfull.com
R Data Mining

Implement data mining techniques through practical

use cases and real-world datasets

Andrea Cirillo

BIRMINGHAM - MUMBAI
R Data Mining
Copyright © 2017 Packt Publishing

All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or
transmitted in any form or by any means, without the prior written permission of the
publisher, except in the case of brief quotations embedded in critical articles or reviews.

Every effort has been made in the preparation of this book to ensure the accuracy of the
information presented. However, the information contained in this book is sold without
warranty, either express or implied. Neither the author, nor Packt Publishing, and its
dealers and distributors will be held liable for any damages caused or alleged to be caused
directly or indirectly by this book.

Packt Publishing has endeavored to provide trademark information about all of the
companies and products mentioned in this book by the appropriate use of capitals.
However, Packt Publishing cannot guarantee the accuracy of this information.

First published: November 2017

Production reference: 1271117

Published by Packt Publishing Ltd.

Livery Place
35 Livery Street
Birmingham
B3 2PB, UK.

ISBN 978-1-78712-446-2

www.packtpub.com
 Credits
Author Copy Editors
Andrea Cirillo Safis Editing
Vikrant Phadkay

Reviewers Project Coordinator

Enrico Pegoraro Nidhi Joshi
Doug Ortiz
Radovan Kavicky
Oleg Okun

Commissioning Editor Proofreader

Amey Varangaonkar Safis Editing

Acquisition Editor Indexer

Varsha Shetty Tejal Daruwale Soni

Content Development Editor Graphics

Mayur Pawanikar Tania Dutta

Technical Editor Production Coordinator

Karan Thakkar Aparna Bhagat
 About the Author
Andrea Cirillo is currently working as an audit quantitative analyst at Intesa Sanpaolo
Banking Group. He gained financial and external audit experience at Deloitte Touche
Tohmatsu and internal audit experience at FNM, a listed Italian company. His main
responsibilities involve the evaluation of credit risk management models and their
enhancement, mainly within the field of the Basel III capital agreement. He is married to
Francesca and is the father of Tommaso, Gianna, Zaccaria, and Filippo. Andrea has written
and contributed to a few useful R packages such as updateR, ramazon, and paletteR, and
regularly shares insightful advice and tutorials on R programming. His research and work
mainly focus on the use of R in the fields of risk management and fraud detection, largely
by modeling custom algorithms and developing interactive applications.

Andrea has previously authored RStudio for R Statistical Computing Cookbook for Packt
Publishing.

To Cesca, Tommaso, Gianna, Zaccaria and Filippo.

 About the Reviewers
Enrico Pegoraro graduated in statistics from the Italian University of Padua more than 20
years ago. He says that "he has experienced in himself the fast-growing computer science and
statistics worlds". He has worked on projects involving databases, software development,
programming languages, data integration, Linux, Windows, and cloud computing. He is
currently working as a freelance statistician and data scientist.

Enrico has gained more than 10 years of experience with R and other statistical software
training and consulting activities, with a special focus on Six Sigma, industrial statistical
analysis, and corporate training courses. He is also a partner of the main company
supporting the MilanoR Italian community. In this company, he works as a freelance
principal data scientist, as well as teacher of statistical models and data mining with R
training courses.

In his first job, Enrico collaborated with Italian medical institutions, contributing to some
regional projects/publications on nosocomial infections. His main expertise is in consulting
and teaching statistical modeling, data mining, data science, medical statistics, predictive
models, SPC, and industrial statistics. Enrico planning to develop an Italian-language
website dedicated to R (www.r-project.it).

Enrico can be contacted at pego.enrico@tiscalil.it.

I would like to thank all the people who support me and my activities, particularly my
partner, Sonja, and her son, Gianluca.
Doug Ortiz is an enterprise cloud, big data, data analytics, and solutions architect who has
been architecting, designing, developing, and integrating enterprise solutions throughout
his career. Organizations that leverage his skillset have been able to rediscover and reuse
their underutilized data via existing and emerging technologies such as Amazon Web
Services, Microsoft Azure, Google Cloud, Microsoft BI Stack, Hadoop, Spark, NoSQL
databases, and SharePoint along with related toolsets and technologies.

He is also the founder of Illustris, LLC and can be reached at dougortiz@illustris.org.

Some interesting aspects of his profession are:

Experience in integrating multiple platforms and products

Big data, data science, R, and Python Certifications
He helps organizations gain a deeper understanding of the value of their current
investments in data and existing resources, turning them into useful sources of
information
He has improved, salvaged, and architected projects by utilizing unique and
innovative techniques
He regularly reviews books on Amazon Web Services, data science, machine
learning, R, and cloud technologies

His hobbies are yoga and scuba diving.

I would like to thank my wonderful wife, Mila, for all her help and support, as well as
Maria, Nikolay, and our wonderful children.
Radovan Kavicky is the principal data scientist and president at GapData Institute, based in
Bratislava, Slovakia, where he harnesses the power of data and wisdom of economics for
public good. He is a macroeconomist by education, and consultant and analyst by
profession (8+ years of experience in consulting for clients from the public and private
sector), with strong mathematical and analytical skills. He is able to deliver top-level
research and analytical work. From MATLAB, SAS, and Stata, he switched to Python, R and
Tableau.

Radovan is an evangelist of open data and a member of the Slovak Economic Association
(SEA), Open Budget Initiative, Open Government Partnership, and the global Tableau
#DataLeader network (2017). He is the founder of PyData Bratislava, R <- Slovakia, and the
SK/CZ Tableau User Group (skczTUG). He has been a speaker at @TechSummit (Bratislava,
2017) and @PyData (Berlin, 2017).

You can follow him on Twitter at @radovankavicky, @GapDataInst or @PyDataBA. His

full profile and experience are available at https:/​/​www.​linkedin.​com/​in/
radovankavicky/​ and https:/​/​github.​com/​radovankavicky.

GapData Institute: https:/​/​www.​gapdata.​org.

Oleg Okun is a machine learning expert and author/editor of four books, numerous journal
articles, and many conference papers. His career spans more than a quarter of a century. He
was employed in both academia and industry in his mother country, Belarus, and abroad
(Finland, Sweden, and Germany). His work experience includes document image analysis,
fingerprint biometrics, bioinformatics, online/offline marketing analytics, credit scoring
analytics, and text analytics.

He is interested in all aspects of distributed machine learning and the Internet of Things.
Oleg currently lives and works in Hamburg, Germany.

I would like to express my deepest gratitude to my parents for everything that they have
done for me.
 www.PacktPub.com
For support files and downloads related to your book, please visit www.PacktPub.com.

Did you know that Packt offers eBook versions of every book published, with PDF and
ePub files available? You can upgrade to the eBook version at www.PacktPub.com and as a
print book customer, you are entitled to a discount on the eBook copy. Get in touch with us
at service@packtpub.com for more details.

At www.PacktPub.com, you can also read a collection of free technical articles, sign up for a
range of free newsletters and receive exclusive discounts and offers on Packt books and
eBooks.

https://www.packtpub.com/mapt

Get the most in-demand software skills with Mapt. Mapt gives you full access to all Packt
books and video courses, as well as industry-leading tools to help you plan your personal
development and advance your career.

Why subscribe?
Fully searchable across every book published by Packt
Copy and paste, print, and bookmark content
On demand and accessible via a web browser
 Customer Feedback
Thanks for purchasing this Packt book. At Packt, quality is at the heart of our editorial
process. To help us improve, please leave us an honest review on this book's Amazon page
at https:/​/​www.​amazon.​com/​dp/​1787124460.

If you'd like to join our team of regular reviewers, you can e-mail us at
customerreviews@packtpub.com. We award our regular reviewers with free eBooks and
videos in exchange for their valuable feedback. Help us be relentless in improving our
products!
Visit https://textbookfull.com
now to explore a rich
collection of eBooks, textbook
and enjoy exciting offers!
 Table of Contents
Preface 1
Chapter 1: Why to Choose R for Your Data Mining and Where to Start 6
What is R? 7
A bit of history 8
R's points of strength 8
Open source inside 8
Plugin ready 9
Data visualization friendly 10
Installing R and writing R code 12
Downloading R 12
R installation for Windows and macOS 13
R installation for Linux OS 14
Main components of a base R installation 14
Possible alternatives to write and run R code 15
RStudio (all OSs) 16
The Jupyter Notebook (all OSs) 18
Visual Studio (Windows users only) 18
R foundational notions 19
A preliminary R session 20
Executing R interactively through the R console 21
Creating an R script 22
Executing an R script 23
Vectors 24
Lists 26
Creating lists 26
Subsetting lists 27
Data frames 27
Functions 29
R's weaknesses and how to overcome them 31
Learning R effectively and minimizing the effort 33
The tidyverse 34
Leveraging the R community to learn R 34
Where to find the R community 35
Engaging with the community to learn R 35
Handling large datasets with R 37
Further references 38
Table of Contents

Summary 38
Chapter 2: A First Primer on Data Mining Analysing Your Bank Account
Data 39
Acquiring and preparing your banking data 40
Data model 40
Summarizing your data with pivot-like tables 41
A gentle introduction to the pipe operator 43
An even more gentle introduction to the dplyr package 45
Installing the necessary packages and loading your data into R 45
Installing and loading the necessary packages 46
Importing your data into R 46
Defining the monthly and daily sum of expenses 47
Visualizing your data with ggplot2 51
Basic data visualization principles 51
Less but better 52
Not every chart is good for your message 54
Scatter plot 54
Line chart 55
Bar plot 55
Other advanced charts 56
Colors have to be chosen carefully 57
A bit of theory - chromatic circle, hue, and luminosity 57
Visualizing your data with ggplot 60
One more gentle introduction – the grammar of graphics 61
A layered grammar of graphics – ggplot2 61
Visualizing your banking movements with ggplot2 62
Visualizing the number of movements per day of the week 62
Further references 66
Summary 66
Chapter 3: The Data Mining Process - CRISP-DM Methodology 68
The Crisp-DM methodology data mining cycle 69
Business understanding 71
Data understanding 71
Data collection 71
How to perform data collection with R 72
Data import from TXT and CSV files 72
Data import from different types of format already structured as tables 72
Data import from unstructured sources 72
Data description 73
How to perform data description with R 73
Data exploration 73
What to use in R to perform this task 74

[ ii ]
Table of Contents

The summary() function 74

Box plot 75
Histograms 77
Data preparation 78
Modelling 79
Defining a data modeling strategy 79
How similar problems were solved in the past 80
Emerging techniques 80
Classification of modeling problems 80
How to perform data modeling with R 81
Evaluation 81
Clustering evaluation 81
Classification evaluation 82
Regression evaluation 83
How to judge the adequacy of a model's performance 84
What to use in R to perform this task 85
Deployment 85
Deployment plan development 86
Maintenance plan development 86
Summary 88
Chapter 4: Keeping the House Clean – The Data Mining Architecture 89
A general overview 90
Data sources 91
Types of data sources 92
Unstructured data sources 93
Structured data sources 93
Key issues of data sources 94
Databases and data warehouses 94
The third wheel – the data mart 95
One-level database 96
Two-level database 96
Three-level database 97
Technologies 97
SQL 98
MongoDB 98
Hadoop 99
The data mining engine 99
The interpreter 100
The interface between the engine and the data warehouse 100
The data mining algorithms 101
User interface 102

[ iii ]
Table of Contents

Clarity 102
Clarity and mystery 103
Clarity and simplicity 103
Efficiency 103
Consistency 104
Syntax highlight 105
Auto-completion 106
How to build a data mining architecture in R 107
Data sources 107
The data warehouse 108
The data mining engine 108
The interface between the engine and the data warehouse 109
The data mining algorithms 109
The user interface 109
Further references 110
Summary 111
Chapter 5: How to Address a Data Mining Problem – Data Cleaning and
Validation 112
On a quiet day 113
Data cleaning 115
Tidy data 115
Analysing the structure of our data 117
The str function 117
The describe function 118
head, tail, and View functions 119
Evaluating your data tidiness 121
Every row is a record 121
Every column shows an attribute 122
Every table represents an observational unit 123
Tidying our data 124
The tidyr package 124
Long versus wide data 124
The spread function 127
The gather function 128
The separate function 129
Applying tidyr to our dataset 130
Validating our data 132
Fitness for use 132
Conformance to standards 133
Data quality controls 133
Consistency checks 134
Data type checks 134
Logical checks 134
Domain checks 135

[ iv ]
Table of Contents

Uniqueness checks 135

Performing data validation on our data 135
Data type checks with str() 135
Domain checks 138
The final touch — data merging 143
left_join function 144
moving beyond left_join 146
Further references 146
Summary 147
Chapter 6: Looking into Your Data Eyes – Exploratory Data Analysis 148
Introducing summary EDA 149
Describing the population distribution 149
Quartiles and Median 149
Mean 151
The mean and phenomenon going on within sub populations 152
The mean being biased by outlier values 154
Computing the mean of our population 155
Variance 155
Standard deviation 156
Skewness 157
Measuring the relationship between variables 160
Correlation 161
The Pearson correlation coefficient 162
Distance correlation 167
Weaknesses of summary EDA - the Anscombe quartet 168
Graphical EDA 169
Visualizing a variable distribution 169
Histogram 170
Reporting date histogram 170
Geographical area histogram 171
Cash flow histogram 171
Boxplot 173
Checking for outliers 174
Visualizing relationships between variables 176
Scatterplots 176
Adding title, subtitle, and caption to the plot 178
Setting axis and legend 179
Adding explicative text to the plot 180
Final touches on colors 181
Further references 182
Summary 182
Chapter 7: Our First Guess – a Linear Regression 184
Defining a data modelling strategy 185
Data modelling notions 190

[v]
Table of Contents

Supervised learning 190

Unsupervised learning 191
The modeling strategy 191
Applying linear regression to our data 192
The intuition behind linear regression 192
The math behind the linear regression 194
Ordinary least squares technique 195
Model requirements – what to look for before applying the model 196
Residuals' uncorrelation 196
Residuals' homoscedasticity 197
How to apply linear regression in R 197
Fitting the linear regression model 198
Validating model assumption 198
Visualizing fitted values 200
Preparing the data for visualization 204
Developing the data visualization 205
Further references 207
Summary 207
Chapter 8: A Gentle Introduction to Model Performance Evaluation 209
Defining model performance 210
Fitting versus interpretability 210
Making predictions with models 212
Measuring performance in regression models 214
Mean squared error 215
R-squared 220
R-squared meaning and interpretation 222
R-squared computation in R 223
Adjusted R-squared 224
R-squared misconceptions 224
The R-squared doesn't measure the goodness of fit 224
A low R-squared doesn't mean your model is not statistically significant 226
Measuring the performance in classification problems 227
The confusion matrix 228
Confusion matrix in R 229
Accuracy 231
How to compute accuracy in R 231
Sensitivity 233
How to compute sensitivity in R 233
Specificity 233
How to compute specificity in R 234
How to choose the right performance statistics 234
A final general warning – training versus test datasets 235
Further references 236

[ vi ]
Table of Contents

Summary 236
Chapter 9: Don't Give up – Power up Your Regression Including
Multiple Variables 238
Moving from simple to multiple linear regression 239
Notation 239
Assumptions 239
Variables' collinearity 240
Tolerance 241
Variance inflation factors 243
Addressing collinearity 243
Dimensionality reduction 243
Stepwise regression 244
Backward stepwise regression 245
From the full model to the n-1 model 246
Forward stepwise regression 248
Double direction stepwise regression 248
Principal component regression 249
Fitting a multiple linear model with R 251
Model fitting 251
Variable assumptions validation 254
Residual assumptions validation 256
Dimensionality reduction 257
Principal component regression 257
Stepwise regression 260
Linear model cheat sheet 265
Further references 266
Summary 266
Chapter 10: A Different Outlook to Problems with Classification Models 267
What is classification and why do we need it? 268
Linear regression limitations for categorical variables 268
Common classification algorithms and models 270
Logistic regression 272
The intuition behind logistic regression 272
The logistic function estimates a response variable enclosed within an upper and
lower bound 273
The logistic function estimates the probability of an observation pertaining to one of
the two available categories 274
The math behind logistic regression 274
Maximum likelihood estimator 276
Model assumptions 277
Absence of multicollinearity between variables 277
Linear relationship between explanatory variables and log odds 278

[ vii ]
Table of Contents

Large enough sample size 279

How to apply logistic regression in R 279
Fitting the model 281
Reading the glm() estimation output 281
The level of statistical significance of the association between the explanatory
variable and the response variable 282
The AIC performance metric 284
Validating model assumptions 285
Fitting quadratic and cubic models to test for linearity of log odds 285
Visualizing and interpreting logistic regression results 287
Visualizing results 287
Interpreting results 290
Logistic regression cheat sheet 292
Support vector machines 292
The intuition behind support vector machines 293
The hyperplane 293
Maximal margin classifier 296
Support vector and support vector machines 296
Model assumptions 297
Independent and identically distributed random variables 297
Independent variables 298
Identically distributed 298
Applying support vector machines in R 301
The svm() function 301
Applying the svm function to our data 302
Interpreting support vector machine results 303
Understanding the meaning of hyperplane weights 303
Support Vector Machine cheat sheet 305
References 306
Summary 306
Chapter 11: The Final Clash – Random Forests and Ensemble Learning 308
Random forest 309
Random forest building blocks – decision trees introduction 309
The intuition behind random forests 313
How to apply random forests in R 314
Evaluating the results of the model 315
Performance of the model 316
OOB estimate error rate 316
Confusion matrix 318
Importance of predictors 319
Mean decrease in accuracy 319
Gini index 319
Plotting relative importance of predictors 320
Random forest cheat sheet 321

[ viii ]
Table of Contents

Ensemble learning 322

Basic ensemble learning techniques 322
Applying ensemble learning to our data in R 323
The R caret package 323
Computing a confusion matrix with the caret package 324
Interpreting confusion matrix results 327
Applying a weighted majority vote to our data 329
Applying estimated models on new data 331
predict.glm() for prediction from the logistic model 333
predict.randomForest() for prediction from random forests 333
predict.svm() for prediction from support vector machines 333
A more structured approach to predictive analytics 334
Applying the majority vote ensemble technique on predicted data 335
Further references 337
Summary 337
Chapter 12: Looking for the Culprit – Text Data Mining with R 339
Extracting data from a PDF file in R 340
Getting a list of documents in a folder 341
Reading PDF files into R via pdf_text() 342
Iteratively extracting text from a set of documents with a for loop 343
Sentiment analysis 348
Developing wordclouds from text 351
Looking for context in text – analyzing document n-grams 353
Performing network analysis on textual data 355
Obtaining an hedge list from a data frame 359
Visualizing a network with the ggraph package 360
Tuning the appearance of nodes and edges 362
Computing the degree parameter in a network to highlight relevant nodes 363
Further references 365
Summary 365
Chapter 13: Sharing Your Stories with Your Stakeholders through R
Markdown 366
Principles of a good data mining report 366
Clearly state the objectives 367
Clearly state assumptions 367
Make the data treatments clear 368
Show consistent data 368
Provide data lineage 369
Set up an rmarkdown report 369

[ ix ]
Table of Contents

Develop an R markdown report in RStudio 372

A brief introduction to markdown 372
Inserting a chunk of code 374
How to show readable tables in rmarkdwon reports 376
Reproducing R code output within text through inline code 377
Introduction to Shiny and the reactivity framework 378
Employing input and output to deal with changes in Shiny app parameters 379
Adding an interactive data lineage module 383
Adding an input panel to an R markdown report 384
Adding a data table to your report 385
Expanding Shiny beyond the basics 387
Rendering and sharing an R markdown report 387
Rendering an R markdown report 387
Sharing an R Markdown report 389
Render a static markdown report into different file formats 390
Render interactive Shiny apps on dedicated servers 390
Sharing a Shiny app through shinyapps.io 391
Further references 392
Summary 392
Chapter 14: Epilogue 393
Chapter 15: Dealing with Dates, Relative Paths and Functions 397
Dealing with dates in R 397
Working directories and relative paths in R 397
Conditional statements 399
Index 400

[x]
Visit https://textbookfull.com
now to explore a rich
collection of eBooks, textbook
and enjoy exciting offers!
 Preface
You have probably heard that R is a fabulous tool that is gaining in popularity everyday
among data analysts and data scientists, and that it is renowned for its ability to deliver
highly flexible and professional results, paired with astonishing data visualizations. All this
sounds great, but how can you learn to use R as a data mining tool? This book will guide
you from the very beginning of this journey; you will not need to bring anything with you
except your curiosity, since we will discover everything we need along the way.

The book will help you develop these powerful skills through immersion in a crime case
that requires the use of data mining skills to solve, where you will be asked to help resolve a
real fraud case affecting a commercial company using both basic and advanced data mining
techniques.

At the end of our trip into the R world, you will be able to identify data mining problems,
analyze them, and correctly address them with the main data mining techniques (and some
advanced ones), producing astonishing final reports to convey messages and narrate the
stories you found within your data.

What this book covers

Chapter 1, Why to Choose R for Your Data Mining and Where to Start, gives you some relevant
facts about R's history, its main strengths and weaknesses, and how to install the language
on your computer and write basic code.

Chapter 2, A First Primer on Data Mining -Analyzing Your Bank Account Data, applies R to
our data.

Chapter 3, The Data Mining Process - the CRISP-DM Methodology, teaches you to organize
and conduct a data mining project through the CRISP-DM methodology.

Chapter 4, Keeping the Home Clean – The Data Mining Architecture, defines the static part of
our data mining projects, the data mining architecture.

Chapter 5, How to Address a Data Mining Problem – Data Cleaning and Validation, covers data
quality and data validation, where you will find out which metrics define the level of
quality of our data and discover a set of checks that can be employed to assess this quality.
Preface

Chapter 6, Looking into Your Data Eyes – Exploratory Data Analysis, teaches you about the
concept of exploratory data analysis and how it can be included within the data analysis
process.

Chapter 7, Our First Guess – A Linear Regression, lets us estimate a simple linear regression
model and check whether its assumptions have been satisfied.

Chapter 8, A Gentle Introduction to Model Performance Evaluation, covers the tools used to
define and measure the performance of data mining models.

Chapter 9, Don't Give Up – Power Up Your Regression Including Multiple Variables, predicts
the output of our response variable when more than one exploratory variable is involved.

Chapter 10, A Different Outlook to Problems with Classification Models, looks into classification
models, the need of them and they are uses.

Chapter 11, The Final Clash – Random Forest and Ensemble Learning, in this chapter we will
learn how to apply ensemble learning to estimated classification models.

Chapter 12, Looking for the Culprit – Text Data Mining with R, shows how to prepare the data
frame for text mining activities, removing irrelevant words and transforming it from a list
of sentences to a list of words. You also learn to perform sentiment analyses, wordcloud
development, and n-gram analyses on it.

Chapter 13, Sharing Your Stories with Your Stakeholders through R Markdown, employs R
markdown and shiny, two powerful instruments made available within the RStudio
ecosystem.

Chapter 14, Epilogue, is the unique background story made to learn the topics in a very
engaging manner.

Appendix, Dealing with Dates, Relative Paths, and Functions, includes additional information
to get things running in R.

What you need for this book

You will easily be able to sail through the chapters by employing R and UNIX or Windows.
The version used is R 3.4.0.

[2]
Preface

Who this book is for

If you are a budding data scientist or a data analyst with basic knowledge of R, and you
want to get into the intricacies of data mining in a practical manner, this is the book for you.
No previous experience of data mining is required.

Conventions
In this book, you will find a number of text styles that distinguish between different kinds
of information. Here are some examples of these styles and an explanation of their meaning.

Code words in text, database table names, folder names, filenames, file extensions, path
names, dummy URLs, user input, and Twitter handles are shown as follows: "Finally,
ggplot2 gives you the ability to highly customize your plot, adding every kind of graphical
or textual annotation to it."

A block of code is set as follows:

install.packages("ggplot2")
library(ggplot2)

New terms and important words are shown in bold.

Words that you see on the screen, for example, in menus or dialog boxes, appear in the text
like this: "In order to download new modules, we will go to Files | Settings | Project Name
| Project Interpreter."

Warnings or important notes appear like this.

Tips and tricks appear like this.

[3]
Preface

Reader feedback
Feedback from our readers is always welcome. Let us know what you think about this
book-what you liked or disliked. Reader feedback is important for us as it helps us develop
titles that you will really get the most out of. To send us general feedback, simply email
feedback@packtpub.com, and mention the book's title in the subject of your message. If
there is a topic that you have expertise in and you are interested in either writing or
contributing to a book, see our author guide at www.packtpub.com/authors.

Customer support
Now that you are the proud owner of a Packt book, we have a number of things to help you
to get the most from your purchase.

Downloading the example code

You can download the example code files for this book from your account at
http://www.packtpub.com. If you purchased this book elsewhere, you can visit
http://www.packtpub.com/support and register to have the files emailed directly to you.
You can download the code files by following these steps:

1. Log in or register to our website using your email address and password.
2. Hover the mouse pointer on the SUPPORT tab at the top.
3. Click on Code Downloads & Errata.
4. Enter the name of the book in the Search box.
5. Select the book for which you're looking to download the code files.
6. Choose from the drop-down menu where you purchased this book from.
7. Click on Code Download.

Once the file is downloaded, please make sure that you unzip or extract the folder using the
latest version of:

WinRAR / 7-Zip for Windows

Zipeg / iZip / UnRarX for Mac
7-Zip / PeaZip for Linux

[4]
Preface

The code bundle for the book is also hosted on GitHub at

https://github.com/PacktPublishing/R-Data-Mining. We also have other code bundles
from our rich catalog of books and videos available at
https://github.com/PacktPublishing/. Check them out!

Downloading the color images of this book

We also provide you with a PDF file that has color images of the screenshots/diagrams used
in this book. The color images will help you better understand the changes in the output.
You can download this file from https:/​/​www.​packtpub.​com/​sites/​default/​files/
downloads/​RDataMining_​ColorImages.​pdf.

Errata
Although we have taken every care to ensure the accuracy of our content, mistakes do
happen. If you find a mistake in one of our books-maybe a mistake in the text or the code-
we would be grateful if you could report this to us. By doing so, you can save other readers
from frustration and help us improve subsequent versions of this book. If you find any
errata, please report them by visiting http://www.packtpub.com/submit-errata, selecting
your book, clicking on the Errata Submission Form link, and entering the details of your
errata. Once your errata are verified, your submission will be accepted and the errata will
be uploaded to our website or added to any list of existing errata under the Errata section of
that title. To view the previously submitted errata, go to
https://www.packtpub.com/books/content/support and enter the name of the book in the
search field. The required information will appear under the Errata section.

Piracy
Piracy of copyrighted material on the internet is an ongoing problem across all media. At
Packt, we take the protection of our copyright and licenses very seriously. If you come
across any illegal copies of our works in any form on the internet, please provide us with
the location address or website name immediately so that we can pursue a remedy. Please
contact us at copyright@packtpub.com with a link to the suspected pirated material. We
appreciate your help in protecting our authors and our ability to bring you valuable
content.

Questions
If you have a problem with any aspect of this book, you can contact us at
questions@packtpub.com, and we will do our best to address the problem.

[5]
Random documents with unrelated
content Scribd suggests to you:
of a revolver, and the fourth provided himself with some cartridges
ready for contingencies. After the completion of these preparations, no
one either moved or spoke. The bear meanwhile, moved steadily
towards us, stopping for a moment at the spot where a piece of bread
had intentionally been placed. Just as he stopped to examine it, three
shots in rapid succession went off, and the bear, hit in the head and
chest, lay dead on the ground. The dogs, being let loose, rushed on
their fallen foe and began to tear his shaggy skin. While we were cutting
the bear up, they sat down and watched us, occasionally dipping their
tongues in the warm red blood and snapping up the morsels which were
thrown to them. The bear we had shot was a female, six feet in length;
and after cutting off the tongue and the best portions for meat, we
continued our march in the teeth of the driving snow. One of our people
had cut his finger badly in dressing the bear, and as the application of
chloride of iron did not suffice to stop the violent bleeding, we were
compelled to halt and erect our tent about six o’clock in the evening.
8. When we set out again on the morning of the 12th (the
thermometer marking -26° F.) all round us was a red undulating waste,
and the driving gusts of snow, which hid from our view the nearest
rocky heights, pricked us as if with countless sharp-pointed darts. Such
drifting snow, although it greatly impedes travelling, cannot be
compared with the tremendous snow-storms I had experienced in
Greenland. The same precursory signs were, however, common to both
—extraordinary refractions, brilliant auroras, perfect calms, and a dull
close atmosphere. In taking down the tent, which was covered with
wreaths of snow, every article which fell in it was at once buried under
its drifting waves. Of all the tests of endurance in Arctic journeys none
exceeds that of continuing the march amid driving snow at a low
temperature. Some of my company who had not been accustomed to
walk in such tremendous weather, in attempting to button on their wind-
screens and nose-bands and fasten up their coats after we had left the
tent, at once had their fingers frozen. Our sail-cloth boots were as hard
as stone, and every one took to stamping to preserve his feet from
frost-bite. Under such circumstances the sledge is not packed with that
precision which is the only preservative against the loss of the various
articles of its contents. To watch against this contingency is the special
business of the man who pushes the sledge from behind. Hurry and
confusion were visible in the bag of provisions being left open. At last
everything was ready: the march began, men and dogs, dragging the
sledge along, all coated with snow and entirely covered except the eyes.
In a momentary lull of the wind, we discovered that our march the day
before had led us far too much to the south, and Cape Tegetthoff now
lay before us directly north. Thither we now directed our steps, and as
the wind still came from the north-west, we struck our sledge sail. As a
consequence of this marching against the wind, which is most severely
felt by the leaders of the team, all, even Klotz, had their noses frost-
bitten. We had much difficulty in persuading him to rub his with snow,
urging that his nose did not belong to himself alone, but that seven
noses and fourteen feet were under the general supervision of the
leader, and that each had a share in this general property.
9. As we came under the land, the violence of the snow-drifting
somewhat abated, and in about two hours a calm set in. Close before us
lay the plateau of Cape Tegetthoff, with its steep precipitous sides. From
its summit a line of basalt rocks descended towards the east, ending in
two columns, each about two hundred feet high. We reached them just
before noon, and the weather being propitious we determined the
latitude by observation and found it to be 80° 6′ N.L. The force of the
tide not being able to raise or burst the bay-ice, the thaw-water of the
spring collects itself on the coast-edge in small lakes. Close under one of
these towers of dark-coloured basalt, we set up our tent; and while our
cook was preparing our dinner of bear’s flesh we lay in the sun under
the rocks in order to dry our clothes, which were coated all over with
ice.
 CAPE TEGETTHOFF.
10. About one o’clock I set off with the Tyrolese to the plateau of
Cape Tegetthoff. Those who remained behind spent their time in rubbing
their feet with snow. Lettis had reserved for us the unpleasant surprise
that his feet had been frost-bitten for three hours, and that he had lost
all feeling in them. We marched for an hour on the snow, which lay in
tender azure-blue shadow under the long line of basalt rocks, and after
climbing for another hour over rosy-coloured masses of snow lying
between crystallized rocks, we reached the highest point of the
undulating plateau. No ascent could be more interesting, made, as it
was, in a country so utterly unknown. Haller and Klotz were born
mountaineers, and during my surveys in Tyrol I had made a hundred
ascents of mountains of 10,000 feet, without the tension of expectation
I now experienced, as I mounted this summit. The ascent was not
without difficulty, and it taxed the extraordinary dexterity of the two
Tyrolese to climb up steep icy precipices in their sail-cloth boots. It was
about three o’clock in the afternoon when we reached the summit; the
temperature had fallen to -30° F. (in the tent the thermometer at the
same time marked -24° F. and in the ship -20° F.). By a barometrical
measurement we found the height to be 2,600 feet. Contrary to
expectation the view from the top proved to be limited. In a northerly
direction, the atmosphere, laden with innumerable ice crystals,
possessed so little transparency that Cape Berghaus, at no distance off,
appeared to be covered with a thick veil, and all distant objects were
enveloped in a dense mist. Fogs lay over the interior to the west, and
banks of reddish vapour covered the icy ocean to the south. Some
narrow strips of open water sparkled in the sun. After making a sketch
of all that could be distinctly seen, and determining the bearings of
some points, we returned to the tent. Here we found Lettis and
Cattarinch engaged in rubbing with snow the hands of Lukinovich, which
had been frost-bitten, while he was occupied in rubbing the feet of
Lettis.
11. Nothing except the wind makes men so sensitive to cold as the
want of exercise. The fall of the temperature had been felt far more by
those who remained behind, than by ourselves. Even the wonderful
beauty of the snow-clad summit bathed in rosy light failed to modify
their severe judgment of Franz-Josef Land. Instead of greeting us with
supper ready at the appointed hour, which he ought to have prepared
without the use of spirit, the bewildered cook was vainly endeavouring
to roast bear’s flesh over smoky chips and sticks, and we got our supper
only after I had served out a bottle of alcohol. We then went to rest in
the common sleeping bag, but soon began to shake with cold, which
threw Pospischill, who took oil twice a day for lung-disease, into a fever.
When I left the tent to look at the thermometers, the mercury in one
had gone down into the bulb and was frozen, and the spirits of wine in
the other showed 41° below zero (C.). Some hot grog, for which a
whole bottle of strong rum was used, put us all right, raising the
temperature of our bodies by one or two degrees. After this refreshment
we all fell into a deep sleep, which was incommoded only by the
increasing dampness of our clothes.
12. We started again about six o’clock on the morning of March 13.
The sun had not risen, the spirit of wine thermometer indicated nearly
44° (C.) below zero, and a piercingly cold breeze met us from the land.
Even on board the ship the temperature at the same time marked 37°
(C.) below zero, a difference to be ascribed to the influence of the land
in lowering the temperature. In Greenland we observed still greater
deviations of this nature, which seem to show that climatical influences
are subject to great variations, even in places which are in close
proximity. Cape Berghaus was our goal. From its summit a general view
of the distribution of the land under 80° N. lat. was reasonably to be
expected. Long before the rise of the sun, the hard snowy plains were
tinted with a pale green reflected light, and the icebergs wore a dull
silvery hue, while their outlines constantly changed and undulated. Our
road was formed from millions of glittering snow crystals, so hard that
the sledge glided over them with difficulty and with a creaking noise,
and after three hours, the exertion of dragging had so exhausted us that
we determined to unload the sledge, and, after melting some snow, to
wet its runners with water. A layer of ice was immediately formed on
them, which greatly facilitated the labour of dragging, till it was rubbed
off. A broad inlet surrounded by picturesque mountains—Nordenskjöld
Fiord—had opened out on our left, and as a large glacier formed the
background of this fiord, we took a westerly direction in order to study
the ice-formation. The heights surrounding this fiord seemed equally as
well fitted as Cape Berghaus for the object we had in view. The further
we penetrated into it, the deeper became the layer of fine powdery
snow which the wind had deposited in this hollow. At noon we reached
the high precipitous termination of Sonklar-Glacier, and pitched our tent
by an iceberg.

MELTING SNOW DURING A HALT NEAR CAPE BERGHAUS.

13. In the afternoon, accompanied by the Tyrolese, I ascended a

mountain—Cape Littrow—whose height, by means of an aneroid
barometer, we ascertained to be 2,500 feet. From its summit we had a
view of the mountains of Hall Island, and of the islands which lay to the
east. Not a breath of wind was stirring, and the atmosphere was clearer
than usual, so that, without suffering in the least degree from cold, I
could work for three hours, first in sketching our surroundings and then
in taking observations. From south-west to north-east the peaks of
distant mountains rose above the summits of those in the foreground.
This view, while it assured us that the land we had named after our
monarch must be of great extent, stimulated our impatience to know its
extent, and the nature and relation of its constituent parts. The
Wüllersdorf Mountains were the extreme limits of what could be known
for the present, and their three peaks glowed in the setting sun above
the dark edges of the terraces of the Sonklar-Glacier, whose broad
terminal front over-hung the frozen bay of Nordenskjöld Fiord. It was
eight o’clock in the evening when we returned to our tent, not, however,
before we had made suitable preparations for the observation of the
movement of the glacier. Sumbu and Torossy were our companions; but
we had to tie them with a rope both in going up and coming down, and
we ourselves only mastered the great steepness of the cone of the
mountain by steps which Klotz, who went on before, hewed with
incomparable dexterity and precision in the ice. During the night the
temperature fell to 46° below zero (C.) (-47° F. in the ship), and I do not
believe that we could have passed through it without the help of grog.
We drank it as we lay close together muffled up in our sleeping bag. It
was boiling hot, and so strong, that under other circumstances it must
have made us incapable of work, yet in spite of the grog, we suffered
much all through the night from cold and our frozen clothes.

CHAPTER V.
THE COLD.

1. THE coldest day we had during this expedition was the 14th of
March. By six o’clock on the morning of that day the Tyrolese and I
stood on the summit of the precipitous face of the Sonklar-Glacier. The
others remained behind to clear the tent of snow, and to bury a small
depôt of provisions in an iceberg which was close at hand. The sun had
not yet risen, though a golden gleam behind the glaciers of Salm Island
indicated his near approach. At last the sun himself appeared, blood-
red, glowing with indistinct outline through the mists, and surrounded
with parhelia, which generally occur when the cold is great. The tops of
the high snowy mountains were first touched with rosy light, which
gradually descended and spread over the icy plains, and the sun like a
ball of fire shone at length clearly through the frosty mist, and
everything around seemed on fire. As the sun even at noon was but a
few degrees above the horizon, this wonderful colouring lasted
throughout the day, and the mountains, whose steepest sides were
covered with a frosty efflorescence, shone like glass in this radiant light.
The alcohol thermometer soon after we came on the glacier fell to 59°
1′ (F.) below zero,[34] and a light breeze blowing from the interior, which
would have been pleasant enough on a March day in Europe, exposed
me, while engaged in the indispensable work of drawing and measuring,
to such danger, that though I worked under the shelter of my Tyrolese
companions as a protection against the cold, I was constantly compelled
to rub my stiffened and benumbed hands with snow. We had taken
some rum with us, and as each took his share, he knelt down and
allowed another to shake it into his mouth, without bringing the metal
cup in contact with his lips. This rum, though it was strong, seemed to
have lost all its strength and fluidity. It tasted like innocent milk, and its
consistence was that of oil. The bread was frozen so hard that we
feared to break our teeth in biting it, and it brought blood as we ate it.
The attempt to smoke a cigar was a punishment rather than an
enjoyment, because the icicles on our beards always put them out, and
when we took them out of our mouths they were frozen. Even the
shortest pipes met the same fate. The instruments I used in surveying
seemed to burn when I touched them, and the medals which my
companions wore on their breasts felt like hot iron.

ON THE SONKLAR-GLACIER.

2. The phenomena of cold which we had the opportunity of observing

during this journey, and which I immediately recorded, will perhaps
justify a short break in my narrative while I attempt to describe them.
The horrors of a Scythian winter are an ancient belief, and it used to be
counted wisdom to shun the zones where men were frozen, as well as
the zones where men were scorched. But it has been assumed, with
great exaggeration, that a hot climate makes men sensual and timid,
while a cold climate renders them virtuous and bold. There is far more
truth in the opinion held by some observers, and especially by Polar
navigators, that cold is depressing in its influence, and enfeebles the
powers of the will. At first it stimulates to action, but this vigour is
quickly followed by torpidity; exertion is soon succeeded by the desire to
rest. Persons exposed to these alternations of increased action and
torpor feel as if they were intoxicated. From the stiffness and trembling
of their jaws they speak with great effort, they display uncertainty in all
their movements and the stupor of somnambulists in their actions and
thoughts. Most of the circumpolar animals escape, as much as they can,
the horrors of the frost: some migrate; others, burying themselves in
holes, sleep throughout the winter. The fish, which are found in the
small pools of sweet water on the land are frozen in when these pools
freeze, and awake to life and movement again only when the pools are
thawed.
3. The human body, with an inner warmth amounting to 95°-100° F.,
is exposed in the wastes of North America and Siberia to frightful cold,
the extremes of which have been noted by many different observers.
Back recorded in Fort Reliance, Jan. 17, 1833, the temperature -67° F.;
Hayes, March 17, 1861, -69° F.; Nevérow, in Jakutzk, Jan. 31, 1838,
-74° F.; Kane, -69° F.; Maclure, Jan. 1853, -73° F.; John Ross, 1831,
-56° F.; and Parry, 1821, -55° F.; while the lowest temperature which
has hitherto been observed in the Alpine countries of Europe is only -24°
F. In consequence of the difficulty of observing the extremes of cold,
lower temperatures than these can scarcely ever have been registered.
4. In order to illustrate the effect of an extraordinarily low
temperature on the human frame, the best point to start from is the
imagination of a man exposed without clothes to its influence. At 37° or
50° (C.) of cold a misty halo would encompass him, the edges of which
would have, under certain circumstances, the colours of the rainbow. It
is evident that the moisture of the body rapidly coming forth and
becoming visible in the cold air would cause this mist, which would
decrease with the heat of the body, and disappear on the death of the
frozen man. The purpose of clothing is to counteract as much as
possible this twofold loss of warmth and moisture, which is the principal
cause of the fearful Arctic thirst. But even clothed men exposed to so
low a temperature present a strange appearance. When they are
dragging a sledge on the march their breath streams forth like smoke,
which is soon transformed into a mass of needles of ice, almost hiding
their mouths from view; and the snow on which they tread steams with
the heat which it receives from the snow beneath. The countless crystals
of ice, which fill the air and reduce the clearness of day to a dull yellow
twilight, make a continual rustling noise; their fall in the form of fine
snow-dust, or their floating as frosty vapour, is the cause of that
penetrating feeling of damp which is so perceptible when the cold is
intense, and which receives accretions from the vapours issuing from
the open places of the sea. Notwithstanding all this, there is an
indescribable dryness in the atmosphere, strongly contrasting with the
feeling of dampness. Heavy clouds are impossible; the heavens are
covered only by mists, through which the sun and the moon,
surrounded by halos, glow blood-red. Falls of snow, as we understand
the expression, altogether cease; the snow crystals, under the influence
of cold, are so minute as to be almost invisible. The land, the real home
and source of cold, acts as the great condenser of vapour, and snow and
moisture of every kind, and lies under a deep covering of frozen snow
till the colour of its walls and precipices reappears in April. The soil, in
the stricter sense of the word, is frozen as hard as iron wherever it
appears through the snow, and the mean temperature of Franz-Josef
Land (about 3° F.) makes it highly probable, that the frost penetrates to
the depth of a thousand feet. Great cold, calm weather, and clear
atmosphere combined, are the characteristics of the interior of Arctic
countries. The nearer we approach the sea, the rarer is this
combination. Light breezes sometimes occur with a temperature 37° (C.)
below zero,[35] but the atmosphere is then less transparent.
5. It is well known that sound is propagated far more freely in Polar
regions than with us. When the cold was great, we could hear
conversations, carried on in the usual tone of voice, distinctly at the
distance of several hundred paces. Parry and Middendorf both assert
that the voice is more audible at a distance in cold weather. The
propagation of sound seems to find less hindrance from the irregular
masses of ice and cushions of snow, than from the curtains of our
woods and the carpets of our vegetation. In the mountainous districts of
Europe many of the characteristics of Polar regions, besides intense
cold, are met with; yet it is a fact, that the report of a gun can scarcely
be heard in those situations. Cold, however, can scarcely be regarded as
the essential condition of this phenomenon; for the propagation of
sound, though in a less striking degree, may be observed even in the
summers there.[36] It would seem rather that the amount of moisture in
the atmosphere has a more decided influence in the production of this
phenomenon.
6. When the snow becomes hard as rock, its surface takes a granular
consistence like sugar. Where it lies with its massive wreaths frozen in
the form of billows, our steps resound, as we walk over them, with the
sound as of a drum. The ice is so hard that it emits a ringing sound;
wood becomes wonderfully hard, splits, and is as difficult to cut as
bone; butter becomes like stone; meat must be split, and mercury may
be fired as a bullet from a gun.[37]
7. If cold thus acts on things without life, how much more must it
influence living organisms and the power of man’s will! Cold lowers the
beat of the pulse, weakens the bodily sensations, diminishes the
capacity of movement and of enduring great fatigue. Of all the senses,
taste and smell most lose their force and pungency, the mucous
membrane being in a constant state of congestion and excessive
secretion. After a time a decrease of muscular power is also perceptible.
If one is exposed suddenly to an excessive degree of cold, involuntarily
one shuts the mouth and breathes through the nose; the cold air seems
at first to pinch and pierce the organs of respiration. The eyelids freeze
even in calm weather, and to prevent their closing we have constantly to
clear them from ice, and the beard alone is less frozen than other parts
of the body, because the breath as it issues from the mouth falls down
as snow. Snow-spectacles are dimmed by the moisture of the eyes, and
when the thermometer falls 37° (C.) below zero they are as opaque as
frost-covered windows. The cold, however, is most painfully felt in the
soles of the feet, when there is a cessation of exercise. Nervous
weakness, torpor, and drowsiness follow, which explains the connection
which is usually found between resting and freezing. The most
important point, in fact, for a sledge party, which has such exertions to
make at a very low temperature, is to stand still as little as possible. The
excessive cold which is felt in the soles of the feet during the noon-day
rest is the main reason why afternoon marches make such a demand on
the moral power. Great cold also alters the character of the excretions,
thickens the blood, and increases the need of nourishment from the
increased expenditure of carbon. And while perspiration ceases entirely,
the secretion of the mucous membranes of the nose and eyes is
permanently increased, and the urine assumes almost a deep red colour.
At first the bowels are much confined, a state which, after continuing for
five and sometimes eight days, passes into diarrhœa. The bleaching of
the beard under these influences is a curious fact.
8. Although theoretically, the fat endure cold better than the lean, in
reality this is often reversed. Somewhat in the same way it might be
argued that the negro would have an advantage over the white man, for
the former as a living black bulb thermometer is more receptive of the
warmer waves of heat. But blackening the face or smearing the body
with grease are experiments which could only be recommended by
those who have never been in a position to try them. The only
protection against cold is clothing carefully chosen, and contrivances to
avoid the condensation of moisture. All articles of dress are made as stiff
as iron by the cold. If one puts off his fur coat and lays it down for a few
minutes on the ground, he cannot put it on again till it be thawed. The
fingers of woollen gloves become as unpliable as if they belonged to
mailed gauntlets, and therefore Arctic travellers, except when engaged
in hunting, prefer to use mittens.
9. Constant precautions are needed against the danger of frost-bite,
and the nose of the Arctic voyager especially becomes a most serious
charge. But no sooner has its safety been secured, than the hands
which have rubbed it with snow are threatened with the same fate. The
ears, however, are well protected from frost by the hood. Frost-bite,
which is caused by the stoppage of blood in the capillaries, evinces itself
by a feeling of numbness, which, if not immediately attended to,
increases to a state of complete rigidity. Slight cases are overcome by
rubbing the part affected with snow. When the cold is excessive, feeling
accompanied with a prickling sensation only returns after rubbing for
hours. Under all circumstances, freezing water with an infusion of
hydrochloric acid is the best means of restoring circulation. When the
frost-bitten member is immersed in this, it is at once overspread with a
coating of ice, but as the temperature of the water slowly rises the
frozen limb is gradually thawed. The longer persons are exposed to a
low temperature, the greater becomes their sensitiveness under it. Their
noses, lips and hands swell, and the skin on those parts becomes like
parchment, cracks, and is most sensitive to pain from the least breath of
wind. In cases of neglected frost-bite, the violet colour of a nose or
hand is perpetuated, in spite of all the efforts made to banish it. Frost-
bites of a more severe character will not yield to mere rubbings with
snow, but should be treated with the kind of cold bath we have
described, continued for some days. The formation of blisters, the
swelling of the parts affected, great sensitiveness and liability to a
recurrence of the malady, are the consequences. In many cases a
sensitiveness to changes of temperature lasts for several years.
Amputation is inevitable in severe and neglected cases. When circulation
has been restored, a mixture of iodine and collodion—10 grains to an
ounce—may, according to the experience of Dr. Kepes, be
advantageously applied to reduce the inflammation which generally
results.
10. It is remarkable that great heat as well as great cold should
generate the great evil—thirst. It is also remarkable how rapidly the
demoralisation produced by thirst extends when any one of the party
begins to show signs of suffering from it. Habit, however, enables men
to struggle against thirst more successfully than against hunger. Many
try to relieve it by using snow; which is especially pernicious when its
temperature falls considerably below the point of liquefaction.
Inflammation of the mouth and tongue, rheumatic pains in the teeth,
diarrhœa, and other mischiefs, are the consequences, whenever a party
incautiously yields to the temptation of such a momentary relief. It is in
fact a mere delusion, because it is impossible to eat as much snow—say
a cubic foot—as would be requisite to furnish an adequate amount of
water. Snow of a temperature of 37° to 50° (C.) below zero feels in the
mouth like hot iron, and does not quench, but increases thirst, by its
inflammatory action on the mucous membranes of the parts it affects.
The Eskimos prefer to endure any amount of thirst rather than eat snow,
and it is only the Tschuktschees who indulge in it as a relish with their
food, which is always eaten cold. Snow-eaters during the march were
regarded by us as weaklings, much in the same way as opium-eaters
are. Catarrhs of every kind are less frequent in Polar expeditions, and
the chills to which we are exposed by passing suddenly from the cold of
the land journey to the warmer temperature of the ship, have no evil
consequences. It deserves to be investigated whether this arises from
the difference of the amount of ozone in the atmosphere of the
respective latitudes.—Now let us return to our journey.
11. After crossing over the Sonklar-Glacier and measuring its slight
inclination of 1° 6′, we climbed an elevation to ascertain the most
promising route for penetrating in a northerly direction; and none
seemed better suited than that which lay over its back, which seemed
free from crevasses. But we looked in vain for the fancied paradise of
the interior, which had existed only in our desire to clothe in glowing
colours the Land, from which we had been so long held back. The true
character, however, of Kaiser Franz-Josef Land, so far as it could be
explored in this and the following sledge expeditions, will be the subject
of the next chapter. The accompanying sketch represents a block of
snow, about the height of a man, at the foot of the Sonklar-Glacier, to
which the winds had given a fanlike shape. In the afternoon, after
inspecting the stakes which we had fixed for measuring the motion of
the glacier, we came back to the tent and began our return march to
Cape Tegetthoff and the ship. A cutting wind compelled us to make
constant efforts against frost-bites. With a heavy creaking noise the
sledge was dragged over the hard snow, and to our reduced strength it
seemed to be laden with a double load. The night is generally the
hardest part of such expeditions, and our camping out during the night
under the cliffs of Cape Tegetthoff was especially bitter. Happy was he
who, exhausted by the labour of dragging, fell asleep at once. As usual,
we dug a deep hole in the snow and loosened it as much as possible, so
that we might profit by its property of being one of the worst conductors
of heat. In a short time the inside of the tent was covered with rime
frost, and we ourselves with ice. The tongue only seemed to recover its
former mobility with those who bewailed their loss of knives, stockings,
gloves—yea, of everything, even their place in the tent. They ate their
portion of bear’s flesh much as if they had been chloroformed, and
dropping asleep in their stiffened icy coat of mail, they were awoke by
its gradual thawing, to reiterate without cessation how cold it was; a
fact which no one present was prepared to dispute. The alcohol
thermometer stood at -56° F. (-48° on board the ship), and when the
warmth produced by the exercise we had taken and by the effects of
supper was gone, the feeling of cold was so intense that it seemed far
more probable that we should be frozen to death than that we should
sleep. The cook therefore received orders to brew some strong grog,
and forthwith six spirit-flames burnt under the kettle filled with snow;
but to make snow of such extreme coldness boil quickly we should have
had to place the kettle over Vesuvius itself in the height of an eruption.

BLOCK OF SNOW.

12. We now slept without stirring a limb, and about five o’clock in the
morning of the 15th of March we started to compass the twenty miles
which lay between us and the ship in one march, without encountering
the suffering of another night’s camping out in the snow. The weather
was as clear as it is possible to be at a temperature of -52° F., and going
along with a light breeze from the north, we made use of our sledge sail
to such advantage that we reached the gentle ascent of the west point
of Wilczek Island after a march of seven hours. We formed a second
depôt of provisions on the summit of a rocky promontory, whence we
discerned with a telescope the masts and yards of the ship lying behind
an iceberg, and our fears and anxieties lest it should have drifted away
in our absence were dissipated by this glad view. Our return to the ship
could no longer be a matter of choice; it had become a necessity. Lettis
had been unable for some days to take any share in the labour of
dragging, and walked along in shoes made of reindeer hide, on account
of his frost-bitten feet. Haller also wore similar shoes to save his swollen
feet; Cattarinch’s face was frost-bitten, and he too suffered from
lameness; Pospischill, who could no longer wear his shrunk-up fur coat,
so suffered from frost-bite in both hands, that I sent him on to the ship,
that he might have the help of the doctor as soon as possible. It was
with much effort that we made the last six hours’ march; and when at
length, stiff with ice, we passed between the hummocks that lay around
the ship, Weyprecht, Brosch, Orel, and eight sailors came to meet us,
who, alarmed at the inability of Pospischill to speak in answer to their
questions, had set out from the ship in order to find us.

THE BURIAL OF KRISCH.

13. As I entered my berth I heard the hard breathing of our poor

comrade Krisch. For more than a week he had lain without
consciousness; yet death had not come to relieve him. On the afternoon
of the 16th of March a sudden cessation of all sound told us that he was
no more! Next day, his body, placed in a coffin, was brought on deck,
and our flag hoisted half-mast high. On the 19th, when the thermometer
was at -13° F., the body was committed to its lonely grave in the far
north. A mournful procession left the ship, with a sledge, on which
rested the coffin covered with a flag and cross, and wended its way to
the nearest elevation on the shore of Wilczek Island. Silently struggling
against the drifting snow, we marched on, dragging our burden through
desolate reaches of snow, till we arrived, after a journey of an hour and
a half, at the point we sought on the island. Here, in a fissure between
basaltic columns, we deposited his earthly remains, filling up the cavity
with stones, which we loosened with much labour, and which the wind,
as we stood there, covered with wreaths of snow. We read the prayer
for the dead over him who had shared in our sufferings and trials, but
who was not destined to return home with us with the news of our
success; and close by the spot, surrounded with every symbol of death
and far from the haunts of men, we raised as our farewell a simple
wooden cross. Our sad and solemn task done, there rose in our hearts
the thought, whether we ourselves should be permitted to return home,
or whether we too should find our resting-place in the unapproachable
wastes of the icy north. The wind blowing over the stiff and stark
elevation where we stood, covered us all with a thick coating of snow,
and caused the appearance of frost-bite in the faces and hands of some
of our party. The decoration of the grave of our comrade with a suitable
inscription was therefore deferred till the weather proved more
favourable. We found considerable difficulty in returning to the ship
through an atmosphere filled with snow.[38]

CHAPTER VI.
A GENERAL DESCRIPTION OF KAISER FRANZ-JOSEF LAND.

In now presenting a general view of those parts of Kaiser Franz-Josef

Land which were explored by us, I must be allowed to anticipate the
order of my narrative which describes the subsequent sledge
expeditions, by which our knowledge of the discovered country was so
considerably enlarged.
1. The country, even in its already ascertained extent, is almost as
large as Spitzbergen, and consists of two main masses—Wilczek Land on
the east, and Zichy Land on the west, between which runs a broad
sound called Austria Sound, extending in a northerly direction from Cape
Frankfort till it forks at the extremity of Crown-Prince Rudolf’s Land, 80°
40′ N. L. One branch of it, a broad arm running to the north-east—
Rawlinson Sound—we traced as far as Cape Buda-Pesth. Wilczek and
Zichy Lands are both intersected by many fiords, and numerous islands
lie off their coasts.
2. A continuous surface of ice extends from the one land to the other.
At the time of our exploration, this expanse was formed of ice, for the
most part not more than a year in growth, but crossed in many places
with fissures and broad barriers of piled-up ice. Throughout its whole
extent we saw many icebergs, which we never did in the Novaya Zemlya
seas; whence it is to be inferred that they sail away in a northerly
direction.[39] Our track lay over this ice-sheet. As long as it remains
unbroken, every fiord might serve as a winter harbour; but if it should
break up, not a single locality suitable to form one presented itself along
the coasts we visited, which had no small indentations.[40]
3. The map of this country, which we present, was designed and
constructed from fifteen observations of latitude, from many
observations made with the azimuth compass, from drawings, and from
a system of triangulation, which, from the nature of the circumstances
under which it was formed,[41] makes no pretensions to absolute
exactitude. The heights of the mountains were determined by the
aneroid barometer. Near the ship a base of 2170·8 metres was
measured by Weyprecht and Orel, and connected trigonometrically with
the nearest promontories. This work of theirs formed the basis of my
surveys.
4. It has always been a principle and a practice with Arctic explorers
to name their discoveries either after the promoters of their special
expeditions, or after their predecessors in the work of discovery. Though
they are never likely to become important to the material interests of
mankind, the naming the lands we discovered after those who promoted
our expedition, was, we considered, the most enduring form by which
we could express our gratitude for their efforts in furtherance of a great
idea. The localities, I may add, were named during the work of
surveying.
 5. As I have had the privilege of visiting all the Arctic lands north of
the Atlantic, I have been able to compare them and observe their
resemblances as well as their differences. West Greenland is a high
uniform glacier-plateau; East Greenland is a magnificent Alpine land with
a comparatively rich vegetation and abundant animal life. How and
where the transition between these opposite characters takes place in
the interior is as yet utterly unknown. We may form some notion of
Spitzbergen and Novaya Zemlya, if we imagine a mountain-range, like
that of the Oetzthal with its glaciers, rising from the level of the sea, if
that level were raised about 9,000 feet. There is more softness,
however, in both these countries than is usual in the regions of the high
north. But Franz-Josef Land has all the severity of the higher Arctic
lands; it appears, especially in spring, to be denuded of life of every
kind. Enormous glaciers extend from the lofty solitudes of the
mountains, which rise in bold conical forms. A covering of dazzling
whiteness is spread over everything. The rows of basaltic columns,
rising tier above tier, stand out as if crystallized. The natural colour of
the rocks was not visible, as is usually the case: even the steepest walls
of rock were covered with ice, the consequence of incessant
precipitation, and of the condensation of the excessive moisture on the
cold faces of the rock. This moisture in a country whose mean annual
temperature is about 3° F., seems to indicate its insular character, for
Greenland and Siberia are both remarkable for the dryness of their cold,
and it was singular that even north winds occasioned a fall of
temperature in Franz-Josef Land. In consequence of their enormous
glaciation, and of the frequent occurrence of plateau forms, the new
lands recalled the characteristic features of West Greenland, in the lower
level of the snow-line common to both, and in their volcanic formation.
Isolated groups of conical mountains and table-lands, which are peculiar
to the basaltic formation, constitute the mountain-system of Franz-Josef
Land; chains of mountains were nowhere seen. These mountain forms
are the results of erosion and denudation; there were no isolated
volcanic cones. The mountains, as a rule, are about 2,000 or 3,000 feet
high, except in the south-west, where they attain the height of about
5,000 feet.
6. The later Arctic expeditions have established the existence of vast
volcanic formations in the high north, and of very recent deposits in
their depressions. In fact, a vast volcanic zone seems to extend from
East Greenland, through Iceland, Jan Mayen and Spitzbergen, to Franz-
Josef Land. The geological features of the latter are at any rate in
harmony with those of North-east Greenland. The tertiary Brown-coal
sandstone of East Greenland is also found in Franz-Josef Land, though
Brown-coal itself is met with only in small beds, which, nevertheless,
may be reckoned among the many indications that the climate of Polar
lands must once have been as genial as the climate of Central Europe at
the present day. The kind of rock which predominates is a crystalline
aggregation called by the Swedes “Hyperstenite” (Hypersthene),
identical with the Dolerite of Greenland; but the Dolerite of Franz-Josef
Land is of a coarser-grained texture, and of a dark yellowish green
colour; according to Professor Tschermak (the Director of the Imperial
Mineralogical Museum at Vienna), it consists of Plagioclase, Augite,
Olivine, titaniferous Iron and ferruginous Chlorite. The mountains of this
system forming table-lands, with precipitous rocky sides, give to the
country we discovered its peculiar physiognomy.
7. The Dolerite of Franz-Josef Land greatly resembles also the Dolerite
of Spitzbergen. After the return of the expedition I saw in London some
photographic views of the mountains of North-East Land, Spitzbergen,
taken by Mr. Leigh-Smith, and I was at once struck with the
resemblance between their forms and those of Franz-Josef Land. I learnt
also from Professor Nordenskjöld, the celebrated explorer of
Spitzbergen, as I passed through Sweden, that the rock of North-East
Land was this same Hyperstenite (Hypersthene). Hence the geological
coincidence of Spitzbergen and Franz-Josef Land would seem to be
established; and this geological affinity, viewed in connection with the
existence of lands more or less known, appears to indicate that groups
of islands will be found in the Arctic seas on the north of Europe, as we
know that such abound in the Arctic seas of North America. Gillis’ Land
and King Karl’s Land are, perhaps, the most easterly islands of the
Spitzbergen group; for it is not probable that these and the lands we
discovered form one continuous uninterrupted whole.
8. Amygdaloids, so common in Greenland, were never found by us in
Franz-Josef Land; and while the rocks in the southern portions of the
country were often aphanitic and so far true basalt, in the north they
were coarse-grained and contained Nepheline. The other rocks consisted
of a whitish quartzose sandstone, with a clayey cement, and of another
finely-grained sandstone, containing small granules of quartz and
greenish-grey particles of chlorite, and also of yellowish finely-laminated
clay slate. Erratics, so far as my opportunities permitted me to judge,
were of rare occurrence; but we found many smaller pieces of petrified
wood, allied to lignite.
9. Some of the islands of the Spitzbergen and Franz-Josef Land group
must be of considerable extent, because they bear enormous glaciers,
which are possible only in extensive countries. Their terminal precipices,
sometimes more than 100 feet high, form generally the coast-lines. The
colour of all the glaciers we visited inclined to grey, we seldom found the
dull green-blue hue; the granules of their ice were extraordinarily large;
there were few crevasses; and the moraines were neither large nor
frequent. Their movement was slow; and the snow-line commences at
about 1,000 feet above the level, whereas on the glaciers of Greenland
and Spitzbergen the like limit is generally 2,000 or even 3,000 feet, and
in these countries also, all below that line is free from snow in summer.
Franz-Josef Land, on the contrary, appears even in summer to be buried
under perpetual snow, interrupted only where precipitous rock occurs.
Almost all the glaciers reach down to the sea. Crevasses, even when the
angle of inclination of the glacier is very great, are much less frequent
than in our Alps, and in every respect the lower glacier regions of Franz-
Josef Land approach the character of the névés of our latitudes. There
only was it possible to determine the thickness of the annual deposits of
snow and ice. In these lower portions, the layers were from a foot to a
foot-and-a-half thick; fine veins, about an inch wide, of blue alternating
with streaks of white ice ran through them, which occurred with peculiar
distinctness at the depth of about a fathom. On the whole, this peculiar
structure of alternating bands or veins was not so distinctly marked as it
is in the glaciers of the Alps, because the alternations of temperature
and of the precipitations are very much less in such high latitudes.
10. The glacier ice of Franz-Josef Land was far less dense than the
glacier ice of East Greenland; whence it appears that movement, as a
factor in the structure of the glacier, predominates in Franz-Josef-Land
more than the factor of regelation. Even at the very end of the glaciers,
granules an inch long are distinctly traceable in its layers, and in the
névé region especially the glacier ice is exceedingly porous. The great
tendency of the climate of Franz-Josef Land to promote glaciation is
manifested in the fact, that all the smaller islands are covered with
glaciers with low rounded tops, so that a section through them would
present a regular defined segment of a circle; hence many ice-streams
descending from the summits of the plateaus spread themselves over
the mountain-slopes and need not to be concentrated in valleys and
hollows in order to become glaciers. Yet many glaciers occur—the
Middendorf Glaciers, for example—whose vertical depth amounts to
many hundred feet. Their fissures and the height of the icebergs show
this. It was unfortunately impossible for us to explore the Dove Glacier,
the largest of all we saw, owing to its great distance from the line of our
route. Evaporation from the surface of the glacier goes on with great
intensity during those summer months when the daylight is continual,
and deep water-courses show that streams of thaw-water then flow over
it.
11. The comparison of the temperature of the air within the crevasses
of the glaciers with the external air, invariably proved, that within the
crevasses the temperature was higher. The traces of liquefaction in the
glacier during winter, arising from the warmth of the earth, could not be
observed, because the sides and under-edge of the glaciers were
inaccessible from the enormous masses of snow, and the icicles of the
terminal arches and precipices could be ascribed only to the freezing of
the thaw-water of the preceding summer.
12. The plasticity of the glaciers was so great, that branches of them,
separated by jutting-out rocks, flowed into each other again at their
base, without showing any considerable crevasses. We could only in a
few cases judge of their movement by direct measurement, and we had
never more than one day to test it. One observation made on the
Sonklar Glacier in the month of March did not seem to support the
notion of the advance of the glaciers; but the repetition of similar
experiments, some weeks later, made on two glaciers on the south of
Austria Sound, gave the mean of two inches as the daily movement. It is
very probable that their movement begins in the Arctic regions
somewhat later than in our latitudes, perhaps at the end of July or
beginning of August, because the period of the greatest liquefaction
then ends, while it is at its minimum in March and the beginning of April.
The signs of glacier-movement were apparent in the detachment of
icebergs in the month of March, but more frequently in the month of
May—as at the Simony Glacier—and in the crashing-in of the ice-sheet
at their base in the month of April—as at the Middendorf Glacier; and
the appearance of “glacier dirt,” where there is no material to furnish a
moraine,—as on the Forbes Glacier—must be regarded as a sign of its
onward movement or lateral extension. The infrequency of moraines
may be explained by the resistance which Dolerite offers to weathering,
and may also be regarded as a sign of the slow movement of the
glaciers. Red snow was seen once only, in the month of May, on the
precipices westward of Cape Brünn. We never met with glacier insects,
although they are common in Greenland; and however diligently I
looked for them I never saw unmistakable traces of the grinding and
polishing of rocks by glacier action.
13. It is well known that the north-east of Greenland as well as
Novaya Zemlya and Siberia are slowly rising from the sea, nay, that all
the northern regions of the globe have for ages participated in this
movement. It was, therefore, exceedingly interesting to observe the
characteristic signs of this upheaval in the terraced beaches, covered
with débris containing organic remains along the coast of Austria Sound.
The ebb and flow, which elevates and breaks up the bay-ice only at the
edge, is to be traced on the shores of Austria Sound by a tidal mark of
two feet.
14. The vegetation was everywhere extremely scanty, crushed, not so
much by the intensity of the cold as by its long continuance, and is far
below the vegetation of Greenland, Spitzbergen, and Novaya Zemlya. It
resembled, not indeed in species but in its general character, the
vegetation of the Alps at an elevation of 9,000 or 10,000 feet, while the
Alpine region corresponding to the vegetation of East Greenland lies a
thousand feet lower. We found neither the stunted birches and willows,
nor the numerous phænogamous plants of East Greenland, Spitzbergen,
and Novaya Zemlya. The rare appearance of soil chiefly contributes to
this extremely sparse vegetation, the detritus of the country resembling
the meagre “dirt” layer on an old moraine, here and there enlivened by
a small patch of green. Although we visited Franz-Josef Land at the
season in which vegetation begins to stir, nowhere could there be seen a
patch of sward, even a few feet square, to recall the features of our
latitudes, although we examined depressions very favourably situated
and free from snow. Some level spots showed patches of thin meagre
grasses of Catabrosa algida (Fries), a few specimens of Saxifraga
oppositifolia and of Silene acaulis, rarely Cerastium alpinum or Papaver
nudicale (L.). Thick, cushion-like tufts of mosses were more frequently
discovered. There were abundance of lichens: Imbricaria stygia
(Acharius), Buellia stigmatea (Körber), Gyrophora anthracina (Wulfen),
Cetraria nivalis (Acharius), Usnea melaxantha (Acharius), Bryopogon
jubatus (Körber), Rhizocarpon geographicum (Körber), Sporastatia Morio
(Körber)—and the Umbilicaria arctica of winter, which we found in
Greenland at an elevation of 7,000 feet. These specifications I owe to
the kindness of Professor Fenzl, director of the Botanical Garden in
Vienna, and of Professor Reichhardt. The museum of this institution
accepted the small collection of plants I was able to bring to Europe. Of
some of these there remained nothing but withered roots, so that it was
impossible to determine their character. Nature in those regions, unable
to deck herself with the colours of plants, produces an imposing effect
by her rigid forms, and in summer by the glare of the ice and snow; and
as there are lands which are stifled by the excess of Nature’s gifts and
blessings, so as even to defy efforts of civilization, here in the high
North another extreme is displayed—absolute barrenness and
nakedness, which render it quite uninhabitable.
15. Drift-wood, chiefly of an old date, we frequently found, but in
small quantities. On the shore of Cape Tyrol, we once saw a log of pine
or larch one foot thick and several feet long, lying a little above the
water-line, and which might have been driven thither by the wind, as
the Tegetthoff was. The fragments of wood we found—the branches on
which showed that they did not come from a ship—were of the pine
genus (Pinus picea, Du Roy), and must have come from the southern
regions of Siberia, as the large broad rings of growth showed.
16. Franz-Josef Land is, as may be supposed, entirely uninhabited,
and we never came on any traces of settlements. It is very questionable
whether Eskimos would have been able to find there the means of
subsistence, and if anywhere most likely on the western side of Wilczek
Island, where an “ice-hole” of considerable extent remained open for a
great part of the year.
 17. In the southern parts it is destitute of every kind of animal life,
with the exception of Polar bears and migratory birds. North of Lat. 81°,
the snow bore numberless fresh tracks of foxes, but though their
footmarks were imprinted on the snow beyond the possibility of mistake,
we never saw one. Once we found their excrements, and on Hohenlohe
Island those of an Arctic hare. The scanty vegetation forbade the
presence of the reindeer and musk-ox. It is not, however, impossible
that there may be reindeer in the more westerly parts of the country,
which we did not visit. The character of that particular region
approximates to that of King Karl’s Land and Spitzbergen, on the
pastures of which herds of these animals live and thrive.
 LIPARIS GELATINOSUS.

18. Of the great marine Mammalia, seals only (Phoca grœnlandica

and Phoca barbata) abounded; although we saw some White Whales.
Walruses we saw twice, but not close to the shore; it is, however,
probable that the absence of open water prevented us from seeing the
walrus nearer the shore, for the character of the sea-bottom would
present no obstacle to its existence.
19. Of fish we saw only the species Liparis gelatinosus (Pallas) and a
kind of cod (Gadus), which were taken with the drag-net.
20. The birds, which we found in the region between Novaya Zemlya
and Franz-Josef Land were of the following species:—the long-tailed
Robber Gull (Lestris, K.); the black Robber Gull without the long tail-
feathers; the Burgomaster Gull (Larus Glaucus, B.); the Ice or Ivory Gull
(Larus eburneus); the Kittiwake (Rissa tridactyla, L.); the Sea-swallow
(Sterna macrura, N.); the Arctic Petrel or Mallemoke (Procellaria
glacialis); Ross’s Gull (Rhotostetia rosea); two species of Auks (Uria arra,
P., and Uria mandtii, L.); the Greenland Dove (Grylle columba, Bp.); the
Rotge (Mergulus alle, V.); the Lumme (Mormon arcticus); the Eider-duck
(Somateria mollisima, L.); the Snowy Owl (Strix nivea); the Iceland Knot
(Tringa canutus); the Snow-bunting (Plectrophanes nivalis, M.). Most of
these occurred also on the coasts of Franz-Josef Land.
21. We can here only allude generally to those forms of animal life
which were taken by the drag-net on the south of Franz-Josef Land, and
brought to Europe in the collection of Dr. Kepes, and of which I made
seventy-two drawings. To Professor Heller, of Innspruck, and Professor
Marenzeller, of Vienna, the expedition is indebted for the naming and
arrangement of those specimens, and while I refer my readers to their
fuller account in the Mittheilungen of the Imperial Academy of Sciences
of Vienna, I limit myself here to a few of the results of their
observations. The investigation of the invertebrate Fauna of the sea
through which we passed was necessarily limited from the moment that
the course of the Tegetthoff ceased to be under our control. We had, in
the first place, no zoologist on board, and from the drifting ship nothing
more could be done than letting down the net almost daily during the
weeks of summer—which Lieutenant Weyprecht did—and dragging it for
some hours. The greater part of the animals so taken were immediately
sketched by me, in order that, in the event of the loss of the original
objects, some sort of representation of the animal world of a region
never before investigated might be preserved. The issue justified a
caution which must always be kept in view in Polar expeditions.
Of the abundant shrimp-family of the Arctic seas there are four
species among the collections we formed, namely:—Hippolyte payeri,
Heller, n. sp., Hippolyte turgida (Kröyer), Hippolyte polaris (Sabine), and
Hippolyte borcalis (Owen). The Hippolyte payeri was found at the depth
of 247 metres, and was of a beautiful pink colour and had blue-black
eyes. There were found besides: Crangou boreas and Pandalus borealis
(Kröyer).

HIPPOLYTE PAYERI.

The group of Amphipoda was, comparatively, largely represented

among the Crustacea of the Arctic waters; we often called these Floh-
krebse—flea-crabs—because many of them used their hind legs to hop
along. Eleven species of this genus were brought home in our
collections; among these were Amathillopsis spinigera, a new species,
Cleïppides quadricuspis, also a new species, both described by Professor
Heller; Acanthozone hystrix (Owen), &c. The group—Isopoda—is
represented by the interesting Munnopsis typica (Sars), the Idothea
sabini (Kröyer), and by a new variety, Paranthura arctica.

HYALONEMA LONGISSIMUM.

Of the group Pycnogonida, our collection contained three varieties, of

which two are new.
 UMBELLULA.

Sponges were common; but we were obliged to leave behind the

specimens of the larger kinds on account of the room they took up.
Among the silicious sponges, those of the genus Hyalonema were the
largest in size, and included the forms described as Hyalonema boreale
(Lovèn), and Hyalonema longissimum (Sars). There was one specimen
of the horny sponge, so rare in those parts. The drag-net often brought
up Actiniæ, Bryareum grandiflorum (Sars), and June 2, 1873, from a
depth of 110 fathoms, a specimen of the extremely rare Umbellula
described by Mytius and Ellis, 1753. Since that date this animal had
been lost sight of, until it was found again by the Swedes—Gladans
expedition 1871—in Baffin’s Bay, and by the Challenger, 1873, between
Portugal and Madeira and between Prince Edward’s Island and
Kerguelen’s Land. It may be assumed that our Umbellula is identical with
the form first described, 1758, by Linnæus as Isis encrinus. I regret to
say that this, the most interesting of all the objects we had collected,
was left behind in the Tegetthoff. The sketch of it made from life will
facilitate a comparison with the forms known in other regions and
variously named.

KORETHRASTES HISPIDUS.

NEPHTHYS LONGISETOSA.

Hydroid polypes, widely distributed in several varieties in the Atlantic

Ocean,—Asteridæ and Ophiuridæ, the Korethrastes hispidus (Wyv.
Thomson), a new variety discovered by the Porcupine expedition
between the Faroe and Shetland islands, Crinoidæ, represented by two
species never before found so far north, and several Holothuriæ, were
also among the acquisitions brought home. Our collection was rich in
Annelides, containing seven-and-twenty varieties found in Greenland
and Spitzbergen. Fourteen varieties of Bryozoa were found, and single
specimens of Turbellaria and Gephyrea.

CHAPTER VII.
THE SECOND SLEDGE EXPEDITION.—AUSTRIA SOUND.

1. The first sledge journey enabled me to draw up a plan for a more

extended expedition towards the north. It was not only a cherished
scheme of my own, but it became also the dominating interest on board
the Tegetthoff, although the other scientific investigations were carried
on uninterruptedly. Weyprecht and Brosch continued with admirable
perseverance the laborious observation of the Magnetic Constants, and
measured on the ice close to the ship a base of 2170·8 metres, which
served for all my trigonometrical surveys. The meteorological
observations also were carried on with the usual regularity.
2. For some days the weather had been bad; its increasingly stormy
character excited our fears, lest the ice should break up and the floe
drift away with the ship. The danger of leaving her, in order to explore
the extent of the new country, increased also with the longer duration of
our proposed second journey. We were convinced, too, that the sea
within a few days had broken up the ice almost as far as Wilczek Island,
and a heavy water-sky was seen in the south at no great distance from
us. Discoveries of importance could only be expected from an expedition
of a month’s duration. But withal the venture must be made, and leaving
the dangers and perils to the chances of the future, I gathered together
the picked men who were to accompany me, to lay before them my
plans. I explained to them my design of penetrating in a northerly
direction as far as possible, and I put before them the danger of our
being cut off from the ship. But while I showed the perils, I stimulated
them also by the hope of reward. If the eighty-first degree of latitude
were reached, I guaranteed to them the sum of £100; if we attained the
eighty-second degree, £250; and I declared that merit, and merit alone,
should regulate the distribution of these sums. In order to make sure of
reticence on the part of my company and thus obviate ill-feeling among