Machine Learning with Spark Nick Pentreath

download

https://textbookfull.com/product/machine-learning-with-spark-
nick-pentreath/

Download more ebook from https://textbookfull.com

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

Mastering Apache Spark 2 x Scale your machine learning

and deep learning systems with SparkML DeepLearning4j
and H2O 2nd Edition Romeo Kienzler

https://textbookfull.com/product/mastering-apache-
spark-2-x-scale-your-machine-learning-and-deep-learning-systems-
with-sparkml-deeplearning4j-and-h2o-2nd-edition-romeo-kienzler/

Predictive Analytics with Microsoft Azure Machine

Learning Barga

https://textbookfull.com/product/predictive-analytics-with-
microsoft-azure-machine-learning-barga/

Machine Learning with TensorFlow 2nd Edition Chris A.

Mattmann

https://textbookfull.com/product/machine-learning-with-
tensorflow-2nd-edition-chris-a-mattmann/

Advanced Analytics with Spark Patterns for Learning

from Data at Scale 2nd Edition Sandy Ryza

https://textbookfull.com/product/advanced-analytics-with-spark-
patterns-for-learning-from-data-at-scale-2nd-edition-sandy-ryza/
Natural Language Processing with Spark NLP Learning to
Understand Text at Scale 1st Edition Alex Thomas

https://textbookfull.com/product/natural-language-processing-
with-spark-nlp-learning-to-understand-text-at-scale-1st-edition-
alex-thomas/

Advanced Data Analytics Using Python: With Machine

Learning, Deep Learning and NLP Examples Mukhopadhyay

https://textbookfull.com/product/advanced-data-analytics-using-
python-with-machine-learning-deep-learning-and-nlp-examples-
mukhopadhyay/

Applied Machine Learning for Health and Fitness: A

Practical Guide to Machine Learning with Deep Vision,
Sensors and IoT Kevin Ashley

https://textbookfull.com/product/applied-machine-learning-for-
health-and-fitness-a-practical-guide-to-machine-learning-with-
deep-vision-sensors-and-iot-kevin-ashley/

Stream Processing with Apache Spark Mastering

Structured Streaming and Spark Streaming 1st Edition
Gerard Maas

https://textbookfull.com/product/stream-processing-with-apache-
spark-mastering-structured-streaming-and-spark-streaming-1st-
edition-gerard-maas/

Hands On Machine Learning with Scikit Learn and

TensorFlow Aurélien Géron

https://textbookfull.com/product/hands-on-machine-learning-with-
scikit-learn-and-tensorflow-aurelien-geron/
Preface
In recent years, the volume of data being collected, stored, and
analyzed has exploded, in particular in relation to activity on
the Web and mobile devices, as well as data from the physical
world collected via sensor networks. While large-scale data
storage, processing, analysis, and modeling were previously the
domain of the largest institutions, such as Google, Yahoo!,
Facebook, Twitter, and Salesforce, increasingly, many
organizations are being faced with the challenge of how to
handle a massive amount of data.

When faced with this quantity of data and the common

requirement to utilize it in real time, human-powered systems
quickly become infeasible. This has led to a rise in so-called big
data and machine learning systems that learn from this data to
make automated decisions.

In answer to the challenge of dealing with ever larger-scale data

without any prohibitive cost, new open source technologies
emerged at companies such as Google, Yahoo!, Amazon, and
Facebook, which aimed at making it easier to handle massive
data volumes by distributing data storage and computation
across a cluster of computers.
The most widespread of these is Apache Hadoop, which made
it significantly easier and cheaper to both store large amounts
of data (via the Hadoop Distributed File System, or HDFS) and
run computations on this data (via Hadoop MapReduce, a
framework to perform computation tasks in parallel across
many nodes in a computer cluster).

However, MapReduce has some important shortcomings,

including high overheads to launch each job and reliance on
storing intermediate data and results of the computation to
disk, both of which make Hadoop relatively ill-suited for use
cases of an iterative or low-latency nature. Apache Spark is a
new framework for distributed computing that is designed
from the ground up to be optimized for low-latency tasks and
to store intermediate data and results in memory, thus
addressing some of the major drawbacks of the Hadoop
framework. Spark provides a clean, functional, and easy-to-
understand API to write applications, and is fully compatible
with the Hadoop ecosystem.

Furthermore, Spark provides native APIs in Scala, Java,

Python, and R. The Scala and Python APIs allow all the benefits
of the Scala or Python language, respectively, to be used
directly in Spark applications, including using the relevant
interpreter for real-time, interactive exploration. Spark itself
now provides a toolkit (Spark MLlib in 1.6 and Spark ML in
2.0) of distributed machine learning and data mining models
that is under heavy development and already contains high-
quality, scalable, and efficient algorithms for many common
machine learning tasks, some of which we will delve into in this
book.

Applying machine learning techniques to massive datasets is

challenging, primarily because most well-known machine
learning algorithms are not designed for parallel architectures.
In many cases, designing such algorithms is not an easy task.
The nature of machine learning models is generally iterative,
hence the strong appeal of Spark for this use case. While there
are many competing frameworks for parallel computing, Spark
is one of the few that combines speed, scalability, in-memory
processing, and fault tolerance with ease of programming and a
flexible, expressive, and powerful API design.

Throughout this book, we will focus on real-world applications

of machine learning technology. While we may briefly delve
into some theoretical aspects of machine learning algorithms
and required maths for machine learning, the book will
generally take a practical, applied approach with a focus on
using examples and code to illustrate how to effectively use the
features of Spark and MLlib, as well as other well-known and
freely available packages for machine learning and data
analysis, to create a useful machine learning system.
What this book covers
Chapter 1, Getting Up and Running with Spark, shows how to
install and set up a local development environment for the
Spark framework, as well as how to create a Spark cluster in the
cloud using Amazon EC2. The Spark programming model and
API will be introduced and a simple Spark application will be
created using Scala, Java, and Python.

Chapter 2, Math for Machine Learning, provides a mathematical

introduction to machine learning. Understanding math and
many of its techniques is important to get a good hold on the
inner workings of the algorithms and to get the best results.

Chapter 3, Designing a Machine Learning System, presents an

example of a real-world use case for a machine learning system.
We will design a high-level architecture for an intelligent
system in Spark based on this illustrative use case.

Chapter 4, Obtaining, Processing, and Preparing Data with

Spark, details how to go about obtaining data for use in a
machine learning system, in particular from various freely and
publicly available sources. We will learn how to process, clean,
and transform the raw data into features that may be used in
machine learning models, using available tools, libraries, and
Spark's functionality.

Chapter 5, Building a Recommendation Engine with Spark,

deals with creating a recommendation model based on the
collaborative filtering approach. This model will be used to
recommend items to a given user, as well as create lists of items
that are similar to a given item. Standard metrics to evaluate
the performance of a recommendation model will be covered
here.

Chapter 6, Building a Classification Model with Spark, details

how to create a model for binary classification, as well as how
to utilize standard performance-evaluation metrics for
classification tasks.

Chapter 7, Building a Regression Model with Spark, shows how

to create a model for regression, extending the classification
model created in Chapter 6, Building a Classification Model
with Spark. Evaluation metrics for the performance of
regression models will be detailed here.

Chapter 8, Building a Clustering Model with Spark, explores

how to create a clustering model and how to use related
evaluation methodologies. You will learn how to analyze and
visualize the clusters that are generated.
Chapter 9, Dimensionality Reduction with Spark, takes us
through methods to extract the underlying structure from, and
reduce the dimensionality of, our data. You will learn some
common dimensionality-reduction techniques and how to
apply and analyze them. You will also see how to use the
resulting data representation as an input to another machine
learning model.

Chapter 10, Advanced Text Processing with Spark, introduces

approaches to deal with large-scale text data, including
techniques for feature extraction from text and dealing with the
very high-dimensional features typical in text data.

Chapter 11, Real-Time Machine Learning with Spark

Streaming, provides an overview of Spark Streaming and how
it fits in with the online and incremental learning approaches
to apply machine learning on data streams.

Chapter 12, Pipeline APIs for Spark ML, provides a uniform set
of APIs that are built on top of Data Frames and help the user
to create and tune machine learning pipelines.
What you need for this
book
Throughout this book, we assume that you have some basic
experience with programming in Scala or Python and have
some basic knowledge of machine learning, statistics, and data
analysis.
Who this book is for
This book is aimed at entry-level to intermediate data
scientists, data analysts, software engineers, and practitioners
involved in machine learning or data mining with an interest in
large-scale machine learning approaches, but who are not
necessarily familiar with Spark. You may have some experience
of statistics or machine learning software (perhaps including
MATLAB, scikit-learn, Mahout, R, Weka, and so on) or
distributed systems (including some exposure to Hadoop).
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, pathnames, dummy URLs, user
input, and Twitter handles are shown as follows: "Spark places
user scripts to run Spark in the bin directory."

A block of code is set as follows:

val conf = new SparkConf()

.setAppName("Test Spark App")
.setMaster("local[4]")
val sc = new SparkContext(conf)

Any command-line input or output is written as follows:

>tar xfvz spark-2.1.0-bin-hadoop2.7.tgz

>cd spark-2.1.0-bin-hadoop2.7
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: "These can be obtained from
the AWS homepage by clicking Account | Security Credentials |
Access Credentials."

Warnings or important notes appear in a box like this.

Tips and tricks appear like this.

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
e-mail 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/a
uthors.
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 e-mailed directly to you.

You can download the code files by following these steps:

1. Log in or register to our website using your e-mail 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

The code bundle for the book is also hosted on GitHub at http
s://github.com/PacktPublishing/Machine-Learning-with-Spark-Second-Edit

ion. 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/MachineLearningwithSparkSecondEdition_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.packtpu

b.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.
Getting Up and Running
with Spark
Apache Spark is a framework for distributed computing; this
framework aims to make it simpler to write programs that run
in parallel across many nodes in a cluster of computers or
virtual machines. It tries to abstract the tasks of resource
scheduling, job submission, execution, tracking, and
communication between nodes as well as the low-level
operations that are inherent in parallel data processing. It also
provides a higher level API to work with distributed data. In
this way, it is similar to other distributed processing
frameworks such as Apache Hadoop; however, the underlying
architecture is somewhat different.

Spark began as a research project at the AMP lab in University

of California, Berkeley (https://amplab.cs.berkeley.edu/projects/spa
rk-lightning-fast-cluster-computing/). The university was focused
on the use case of distributed machine learning algorithms.
Hence, it is designed from the ground up for high performance
in applications of an iterative nature, where the same data is
accessed multiple times. This performance is achieved
primarily through caching datasets in memory combined with
low latency and overhead to launch parallel computation tasks.
Together with other features such as fault tolerance, flexible
distributed-memory data structures, and a powerful functional
API, Spark has proved to be broadly useful for a wide range of
large-scale data processing tasks, over and above machine
learning and iterative analytics.

For more information, you can visit:

http://spark.apache.org/community.html

http://spark.apache.org/community.html#history

Performance wise, Spark is much faster than Hadoop for

related workloads. Refer to the following graph:

Source: https://amplab.cs.berkeley.edu/wp-content/uploads/2011/11/spark-lr.png

Spark runs in four modes:

The standalone local mode, where all Spark processes are run
within the same Java Virtual Machine (JVM) process

The standalone cluster mode, using Spark's own built-in, job-

scheduling framework
 Using Mesos, a popular open source cluster-computing
framework

Using YARN (commonly referred to as NextGen

MapReduce), Hadoop

In this chapter, we will do the following:

Download the Spark binaries and set up a development

environment that runs in Spark's standalone local mode. This
environment will be used throughout the book to run the
example code.

Explore Spark's programming model and API using Spark's

interactive console.

Write our first Spark program in Scala, Java, R, and Python.

Set up a Spark cluster using Amazon's Elastic Cloud

Compute (EC2) platform, which can be used for large-sized
data and heavier computational requirements, rather than
running in the local mode.

Set up a Spark Cluster using Amazon Elastic Map Reduce

If you have previous experience in setting up Spark and are

familiar with the basics of writing a Spark program, feel free to
skip this chapter.
Installing and setting up
Spark locally
Spark can be run using the built-in standalone cluster
scheduler in the local mode. This means that all the Spark
processes are run within the same JVM-effectively, a single,
multithreaded instance of Spark. The local mode is very
used for prototyping, development, debugging, and testing.
However, this mode can also be useful in real-world scenarios
to perform parallel computation across multiple cores on a
single computer.

As Spark's local mode is fully compatible with the cluster

mode; programs written and tested locally can be run on a
cluster with just a few additional steps.

The first step in setting up Spark locally is to download the

latest version http://spark.apache.org/downloads.html, which
contains links to download various versions of Spark as well as
to obtain the latest source code via GitHub.

The documents/docs available at http://spark.apache.or

g/docs/latest/ are a comprehensive resource to learn
more about Spark. We highly recommend that you
explore it!
Spark needs to be built against a specific version of Hadoop in
order to access Hadoop Distributed File System (HDFS)
as well as standard and custom Hadoop input sources
Cloudera's Hadoop Distribution, MapR's Hadoop distribution,
and Hadoop 2 (YARN). Unless you wish to build Spark against
a specific Hadoop version, we recommend that you download
the prebuilt Hadoop 2.7 package from an Apache mirror from h
ttp://d3kbcqa49mib13.cloudfront.net/spark-2.0.2-bin-hadoop2.7.tgz.

Spark requires the Scala programming language (version 2.10.x

or 2.11.x at the time of writing this book) in order to run.
Fortunately, the prebuilt binary package comes with the Scala
runtime packages included, so you don't need to install Scala
separately in order to get started. However, you will need to
have a Java Runtime Environment (JRE) or Java
Development Kit (JDK).

Refer to the software and hardware list in this book's

code bundle for installation instructions. R 3.1+ is
needed.

Once you have downloaded the Spark binary package, unpack

the contents of the package and change it to the newly created
directory by running the following commands:

$ tar xfvz spark-2.0.0-bin-hadoop2.7.tgz

$ cd spark-2.0.0-bin-hadoop2.7
Spark places user scripts to run Spark in the bin directory. You
can test whether everything is working correctly by running
one of the example programs included in Spark. Run the
following command:

$ bin/run-example SparkPi 100

This will run the example in Spark's local standalone mode. In

this mode, all the Spark processes are run within the same
JVM, and Spark uses multiple threads for parallel processing.
By default, the preceding example uses a number of threads
equal to the number of cores available on your system. Once
the program is executed, you should see something similar to
the following lines toward the end of the output:

...
16/11/24 14:41:58 INFO Executor: Finished task 99.0 in stage 0.0
(TID 99). 872 bytes result sent to driver
16/11/24 14:41:58 INFO TaskSetManager: Finished task 99.0 in stage
0.0 (TID 99) in 59 ms on localhost (100/100)
16/11/24 14:41:58 INFO DAGScheduler: ResultStage 0 (reduce at
SparkPi.scala:38) finished in 1.988 s
16/11/24 14:41:58 INFO TaskSchedulerImpl: Removed TaskSet 0.0,
whose tasks have all completed, from pool
16/11/24 14:41:58 INFO DAGScheduler: Job 0 finished: reduce at
SparkPi.scala:38, took 2.235920 s
Pi is roughly 3.1409527140952713
The preceding command calls class
org.apache.spark.examples.SparkPi class.

This class takes parameter in the local[N] form, where N is the

number of threads to use. For example, to use only two threads,
run the following command instead:N is the number of threads
to use. Giving local[*] will use all of the cores on the local
machine--that is a common usage.

To use only two threads, run the following command instead:

$ ./bin/spark-submit --class org.apache.spark.examples.SparkPi

--master local[2] ./examples/jars/spark-examples_2.11-2.0.0.jar
Spark clusters
A Spark cluster is made up of two types of processes: a driver
program and multiple executors. In the local mode, all these
processes are run within the same JVM. In a cluster, these
processes are usually run on separate nodes.

For example, a typical cluster that runs in Spark's standalone

mode (that is, using Spark's built-in cluster management
modules) will have the following:

A master node that runs the Spark standalone master process

as well as the driver program

A number of worker nodes, each running an executor process

While we will be using Spark's local standalone mode

throughout this book to illustrate concepts and examples, the
same Spark code that we write can be run on a Spark cluster. In
the preceding example, if we run the code on a Spark
standalone cluster, we could simply pass in the URL for the
master node, as follows:

$ MASTER=spark://IP:PORT --class org.apache.spark.examples.SparkP

./examples/jars/spark-examples_2.11-2.0.0.jar 100
Discovering Diverse Content Through
Random Scribd Documents
Alaskan waters, guarding the seal fisheries and often giving much
needed relief to the whaling fleet that yearly sails from San Francisco
for a cruise in the waters above the Behring Sea.
Officers and crews of the cutters doing service in the waters of
Alaska have remarkable stories to tell, and the log-books of the
cutters Corwin and Bear have been filled during the last twenty-five
years with a record all too brief, of many thrilling adventures in the
frozen North. The Corwin left San Francisco for the Polar Sea in May,
1881, charged with ascertaining, if possible, the fate of two missing
whalers, and to establish communication with the exploring steamer
Jeanette. Five times during the previous year the Corwin had
attempted to reach Herald Island, and failed each time. On this
voyage better success attended, and after braving the perils of the
drift ice, a landing was made, while at the same time the bleak coast
of Wrangel Land was sighted to the westward. On August 12, 1881,
the Corwin having pushed its way through great masses of floating
and grounded ice, into an open space near the island, effected a
landing on Wrangel Land, this being the first time that white men
had ever succeeded in reaching that remote corner of the Arctic
waste.
The cruises of the Corwin in 1880 and 1881 covered over 12,000
miles, and the officers and crew, while carefully preventing illegal
raids upon the sealing interests, also found time to prosecute
important surveys and soundings, to make a careful study of the
natives of Alaska, and to collect a great mass of important data
relative to the natural features and mineral wealth of the country.
The cruises of the Corwin in the succeeding years of 1882, 1883,
1884 and 1885, were of scarcely less importance. One of these
cruises was to St. Lawrence Bay, Siberia, where timely succor was
given to the officers and crew of the burned naval relief steamer
Rogers, which had gone north in the spring of 1881 in search of the
Jeanette. During the Corwin's cruise in 1883 a considerable portion
of the interior of Alaska was carefully explored and an outbreak
among the natives on the mainland promptly quelled. During its two
succeeding cruises the Corwin saved from death nearly 100
shipwrecked whalers and destitute miners.
Since 1885 the cutter Bear has patrolled the Alaskan waters, making
a record equal to that of its predecessor. Its work in protecting the
sealing fisheries is well known, and it has also suppressed in large
measure the illegal sale to the natives of firearms and spirits. Its
record as a life saver is also a long one, and some of its experiences
have been more thrilling than those to be found in the pages of any
romance.
When the Bear reached Alaskan waters in 1887 the captain of the
whaling ship Hunter handed its commander a most remarkable
message, which had been delivered to him a few days before by the
natives of Cape Behring. This message consisted of a piece of wood,
on one side of which was rudely carved: "1887 J. B. V. Bk. Nap.
Tobacco give," and on the other "S. W. C. Nav. M 10 help come."
The riddle offered by the message was speedily solved by the
officers of the Bear. The bark Napoleon had been wrecked in 1885
off Cape Navarin, and only fourteen of the crew of thirty-six men
had been rescued. Of the unlucky twenty-two a few reached the
Siberian shore, but nothing had been heard of their subsequent fate.
The officers of the Bear reasoned that the sender of the message
was a member of the Napoleon's crew who had found refuge with
the natives to the southwest of Cape Navarin and was now anxiously
awaiting rescue. This reasoning proved correct, and a few weeks
later the weary two years' exile of James B. Vincent, of Edgartown,
Mass., boatswain of the Napoleon, had a happy termination.
The story Vincent told his rescuers, was of tragic and absorbing
interest. The Napoleon, caught in a storm, had been wedged in the
ice and its crew compelled to take to the boats. The boats, four in
number, were soon separated, and thirty-six days of fearful suffering
passed before the one containing Vincent and his companions
reached shore. In the meantime nine of the eighteen men in the
boat had died and several others had been driven insane by their
sufferings. Vincent was the only one who could walk when they
reached land. Five more soon died and three of the survivors were
helpless from frost bites and exhaustion when they fell in with a
party of natives. A portion of the latter lived inland, and these took
Vincent with them when they returned to their homes. The following
Spring when the natives visited the shore to fish, Vincent found his
three shipmates barely alive, and they died soon after.
When the fishing was over Vincent went back to the mountains with
his new-found friends, and during the following winter carved and
entrusted to wandering natives from Cape Behring the message
which later brought about his rescue. When spring of the second
year opened Vincent, with the natives, again started for the
seashore to fish. Great was his joy a few weeks later when he was
attracted by the shouting of the natives and looked up to see a
white man and to find himself rescued at last. The Bear conveyed
him to San Francisco, whence he made his way to his home in
Massachusetts.
While among the Eskimo, Vincent was kindly cared for by an old
native, whose wife received him as her son. After a year the
husband died, but his last instructions to his wife were to care for
and keep their guest until he was rescued. When relief at last came
the old woman with tears in her eyes, said that she was ready to
die, for she had done as her husband wished. Warm and tender
hearts can be found even in Siberian wastes.
The Revenue Cutter Service is part of the Treasury Department, and
comes under the direct jurisdiction of the Secretary of the Treasury.
Subordinate to him are a chief and assistant chief of division. Each
vessel of the service patrols the district to which it is assigned, and
forms a picket line at the outer edge of government jurisdiction,
which extends four leagues from the coast. Every vessel arriving in
United States waters is boarded and examined, and its papers
certified. If a vessel liable to seizure or examination does not bring
to when requested to do so, the commander of a cutter, after
discharging a warning gun, has authority to fire into such a vessel,
and all acting under his orders are indemnified from any penalties or
action from damages. On each cutter there are a captain, three
lieutenants, a cadet, an engineer and two assistants, and a crew of a
dozen or more men.
The service includes in its several grades about one thousand men.
Strict discipline is maintained, and its crews receive constant
instruction and exercise in the use of great guns, rifles, carbines,
pistols, cutlasses and the like. An officer of the Revenue Cutter
Service must not only possess considerable executive ability, but
must also be a man of varied and accurate information, having a
knowledge of gunnery and military drills, and be thoroughly familiar
with the customs and navigation laws of the country.
Rank is obtained by promotion, the latter being governed by written
competitive examinations, from three to five of the senior officers of
a lower grade being selected for any vacancy occurring in the higher
grade. A young man wishing to join the service as an officer
undergoes a rigid examination held annually at Washington, and
then serves for several years aboard the revenue schoolship, where
he learns sea mathematics, sea law and seamanship. His period of
apprenticeship ended, he joins a regular cutter as a junior officer
and waits for promotion at a salary of $85 per month.
Life on board a revenue cutter during the months of summer is
usually an easy and pleasant one, but in the winter there is another
and different story to tell. From December to April of each year the
cutters cruise constantly on their stations to give aid to vessels in
distress, and are, in most cases, forbidden to put into port unless
under stress of weather or other unforeseen conditions arise.
Few stormy winter days pass without the revenue cutter seeing a
signal from some vessel in distress, and aid is never sought in vain.
The cutter steers straight for the signal as soon as it is sighted, and
when a quarter of a mile distant lowers a boat. Often a boat is
launched into a sea where death seems certain, but officers and
men never shrink from their duty. When the boat gains the side of
the vessel seeking aid, the master whom misfortune has overtaken,
requests, as a rule, to be towed into port. When such a request is
made, a line must be got to the distressed vessel and from the boat
to the cutter, a task often performed with infinite difficulty and at the
risk of life and limb.
When a vessel is found drifting helplessly and about to dash itself
upon rocks, the peril is even greater. Then the cutter must stand
further away, and its boat is in constant danger of being dashed
upon the rocks. But, thanks to the skill, experience and coolness of
the officers and crew of the cutter, a line is generally got into the
boat and to the steamer, and the imperilled vessel hauled away to
safety.
One of the finest feats of life-saving ever performed by the Revenue
Cutter Service was that credited to the cutter Dexter, some years
ago. On January 17, 1884, the iron-built steamer City of Columbus
left Boston for the port of Savannah, carrying eighty-one passengers
and a ship's company of forty-five persons. Her commander was a
capable and experienced seaman, and though by nightfall the wind,
which had been blowing all day, had increased to a hurricane, and a
heavy sea was running, he had no serious apprehension of danger.
The vessel, following her usual course through Vineyard Sound, had
left behind nearly all the dangerous points which thickly bestrew
those waters, and would soon be safely in the open ocean. It was at
that luckless moment that the captain left the bridge and went
below, first directing the helmsman how to steer.
Within an hour the steamer struck on Devil's Bridge, and an awful
fate was upon the hapless passengers and crew, who were sleeping
soundly, all unconscious of danger. The weather was bitter cold, the
darkness intense, the wind blowing a hurricane and the waves
rolling mountain high. In the twinkling of an eye a hundred poor
creatures were swept to their death in the icy waters. A few of the
stronger ones took refuge in the rigging, but many of these,
benumbed by the cold, dropped one by one from their supports and
disappeared in the sea, while such boats as were cleared away were
either dashed to pieces or instantly swamped.
The wreck occurred about four o'clock in the morning, and soon
after daylight the Dexter reached the scene of the disaster. Her
commander at once dispatched two boats to the rescue of those still
clinging to the rigging of the Columbus, and thirteen men, jumping
from their refuge into the sea, were picked up as they came to the
surface, and conveyed to the Dexter. To reach the wreck in small
boats through an angry sea was an undertaking so perilous as to
make even the boldest pause, and called for courage of the highest
order. However, the Dexter's crew proved equal to the test, and
Lieutenant John U. Rhodes made himself famous by an act of the
noblest heroism. Two men, rendered helpless by cold and exposure,
still clung to the rigging of the Columbus after all their companions
had been taken off. To board the ill-fated vessel was impossible;
Rhodes essayed to reach it by swimming. He gained the side of the
vessel after a gallant battle with the waves, but was struck by a
piece of floating timber, and had to abandon the attempt. Bruised
and half fainting, he insisted upon making another trial, reached the
vessel and brought away the two men, both of whom died a few
hours later. The Legislature of Connecticut, Rhodes' native State,
passed a resolution thanking him for his gallant conduct, and he
received many medals and testimonials.
Rhodes has since died, but the Revenue Cutter Service still numbers
among its officers scores of men endowed with the flawless bravery
of which he gave such shining proof at the wreck of the City of
Columbus. One of these is Lieutenant James H. Scott. This brilliant
young officer—I cite his case as a typical one—was born in
Pennsylvania thirty-seven years ago, and while still in his teens
shipped as a boy on a merchant vessel in commerce between
Philadelphia and Antwerp. Tiring of this trade, he sailed as an able
seaman from New York to Bombay and other East Indian ports,
making the last voyage as boatswain of the good ship Ridgeway,
after which, declining proffer of a second mate's berth, he entered
the Revenue Cutter Service as a cadet.
Graduated in 1890, and made acting third lieutenant on the cutter
Woodbury, it was then that young Scott, who while attached to the
revenue schoolship had jumped overboard in Lisbon harbor and
rescued the quartermaster of his vessel, again gave proof of the
sterling stuff that was in him. On a cold, clear day in January, 1891,
the Woodbury, which is stationed at Portland, Me., was cruising to
the eastward of that port, the thermometer below zero, and the
rigging covered with ice. The Woodbury was about half-way over her
cruising ground when the officer of the deck discovered a large
three-masted schooner hard aground on a ledge of rocks which
stood well out from the shore. A high sea was running at the time,
though the cutter rose and fell to every wave with apparent
unconcern, and breaking clean over the schooner, the crew of which
had taken refuge on the rocks and were now frantically signalling for
help. It was clear that unless help reached them they would quickly
perish from the cold.
Captain Fengar, commanding the Woodbury, ran in as close as he
could without peril to his vessel, and carefully surveyed the ground
before giving an order. His practiced eye told him in a moment that
to send in a boat of the cutter type would mean its certain
destruction against the rocks, even if it could live in the sea then
running. However, the captain suddenly recalled that a fisherman's
village was only a few miles distant, and that there he could obtain a
couple of dories admirably adapted to the task in hand. Shouting to
the men on the rocks to hold on and not lose hope, the cutter, at a
word from its commander, headed about, and went plunging and
rolling at top speed in the direction of the village. Two hours later
the Woodbury was again on the scene, with a good-sized dory on
one of her davits.
Closing in on the wreck, Captain Fengar called for volunteers. Almost
to a man the crew responded, but among the foremost were Cadets
Scott and W. S. Van Cott. Captain Fengar allowed the two young
men to go, but not without some misgivings. Both insisted on pulling
oars, the dory being in charge of Lieutenant W. L. Howland, an
experienced and capable officer. As the dory left the ship it was
observed that a life-saving crew from a station well down the coast
was approaching. It would never do to let the Woodbury be beaten,
and her dory crew pulled with all the vim they could command. The
race was to be a close one, but at the outset the Woodbury's boat
gained the lead, and such a run, in such a sea, was never perhaps
pulled by opposing boats.
Lieutenant Howland in getting close in, dared not run up too close to
the rocks, and after a couple of ineffectual attempts to heave a line
was about to despair of success, when suddenly Cadet (now
Lieutenant) Scott, securing the line around his waist, sprang
overboard, before any one in the boat knew what he was about.
Shouting to Lieutenant Howland to pay the line out, young Scott was
dashed upon the rocks and seized by the imprisoned sailors. The
brave young fellow was badly stunned, but he had gained his point
by getting the line to the rocks. Communication was now effected
with the dory, which all this time was riding the seas at a safe
distance. Another line was hauled up from the boat, and one by one
the sailors jumped clear of the rocks and were hauled to the dory,
whence they were conveyed without delay to the deck of the cutter.
When rescued they had been fourteen hours on the rock. Since the
incident just related, Lieutenant Scott, though still one of its
youngest officers, has held every position in the Revenue Cutter
Service.
The present chief of the Revenue Cutter Service is Captain C. F.
Shoemaker. He has climbed to this position from the lowest rung of
the ladder, and is a man whose success would have been notable in
almost any calling. Many of the other captains of the service are
men of mark and achievement, for the Government has no nobler,
better, braver servants than those who officer and man its revenue
cutters.
 CHAPTER VI
THE OCEAN PILOT

The ocean pilots and deep sea divers of New York have one thing in
common; both object to taking apprentices, and in the case of the
former, at least, there is good reason for this, since they have been,
for generations, the aristocrats of their calling. The pilots who sail
out of Sandy Hook are no hardier than their rugged and fearless
fellows of the North Sea, but they subject themselves to greater
dangers by their long cruises, and rough, indeed, must be the
weather that can keep them in port. They cruise night and day, in
search of incoming craft; their torches' flare lights up the snow and
sleet of winter storms and contends with the darkness of summer
fogs; and they speak and board in all sorts of weather and at all
seasons the fleet liners that cross the western ocean in less than a
week. And these pilots of the New York and New Jersey shores are a
revelation to the tourist, who, having never heard of them, sees
them for the first time. The latter, in most cases, expects to watch a
rough-and-ready sort of fellow in homespun, with a swaggering air
and a boisterous manner, climb from the pilot's yawl up the black
hull's towering side. Instead, he sees a man of modest and pleasing
address, about whom there is little to indicate his calling, and much
that bespeaks the merchant or clerk one meets of a morning on
lower Broadway. There was a time when our pilots indulged in the
luxury of a high silk hat when boarding vessels in sunny weather, but
they are not so fastidious nowadays, and use derbies instead.
Prosperous as a class, the pilots of New York pay dearly for their
prosperity by the most arduous sea labor. Since 1853 more than
thirty-five boats have been sunk and wrecked in various ways, and
twice that number of pilots have lost their lives. There are at the
present time upward of 160 pilots cruising from the port of New
York. They are subject to the supervision of a pilots' commission of
five members, named by the Governor of New York, and each pilot is
appointed after a long and severe apprenticeship. He must first
serve, boy and man, before the mast until he masters every problem
in the management of every form of rig. Then he must contrive to
obtain the position of boat-keeper or pilot's mate. In that capacity he
must serve three full years before he can be admitted for his
examination for a license. After this he must pass a most rigid
examination on all points of seamanship and navigation before the
Board of Pilot Commissioners, and show complete and exact
knowledge of the tides, rips and sands and all other phenomena for
many miles out from the piers of the East and North Rivers.
But even after the candidate has received his license, he is
sometimes forced to wait years, until some pilot happens to die and
leave a vacancy for him. The first year of pilotage he is granted a
license to pilot vessels drawing less than sixteen feet. If he gives
satisfaction, the following year he is permitted to take charge of
vessels drawing eighteen feet. If he passes a satisfactory
examination the third year, he then receives a full license, entitling
him to pilot vessels of any draught, and is then first called a branch
or full pilot. On receiving his license, the pilot must give bonds for
the proper discharge of his duty, and he is liable to heavy fines if he
declines to fill a vacancy or board a vessel making signals for a pilot.
Pilots are paid for their work by the foot, the charges varying
according to the draught. For a ship drawing from twenty-one to
twenty-eight feet they receive $4.88 a foot, and for one drawing six
to thirteen and one-half feet $2.78 a foot, these rates being slightly
increased in winter.
A cruise on a New York pilot-boat, however brief, is an experience
sure to be remembered. When a pilot-boat starts out on a hunt for
ships, it is decided in what order its half-dozen pilots shall take the
prizes, and the man who is to board the first one is placed in
command. The other pilots, meanwhile, take their ease as best suits
their taste, the seaman's work being done by a crew of sailors hired
for the purpose. One pilot, however, is always on the lookout for
sails, and a landsman is compelled to marvel at the certainty with
which these ocean scouts discharge the task of sighting vessels, for
often they are able to tell the name of a steamship before
unaccustomed eyes can discern aught but a waste of waters and a
wide expanse of sky. Still, a part of this skill may be due to the fact
that pilots are always posted before going out as to what vessels are
expected, and from what direction they are coming, the watch being
made all the keener by the fact that the bigger the ship the bigger is
the pilot's pay. A ship, moreover, must take a pilot going out from
the same boat that furnishes the pilot going into port, while if a
captain refuses a pilot he must pay full pilotage, and thus contribute
his tithe to the support of the system. This latter rule seems, at first
glance, a curious provision, but it is defended on the ground that
without it the business would not be remunerative enough for really
competent men to engage in it, and that with unskilled pilots the
annual losses would be greatly in excess of what they are at
present.
When a ship is sighted by daylight, a long blue burgee is hoisted to
the peak of the pilot-boat, which means, "Do you want a pilot?" If
there is no responsive signal, it is taken for granted that the answer
is "Yes," but if a jack is hoisted the watchers know that the vessel
has already been boarded by a pilot from some boat that has sailed
farther away from port in the hunt for a ship. When a ship is sighted
at night she is signalled by means of a torch charged with benzine
and giving forth an intense light. Seen from the other vessel the
effect is startling, the white light illuminating every sail and spar of
the pilot-boat, so that it stands out, its number clearly visible upon
the mainsail, a gray specter against the night's background.
Should the answering signal be favorable, there follows a scene of
great excitement on the deck of the pilot-boat. At first sight of the
ship, the pilot due to take the prize dives down to the cabin, sheds
his working clothes and dons a suit of sober black, and by the time it
is known he is wanted, he is ready to be transferred to his charge.
Taking on a pilot is not without its perils. The yawl nearly always
pitches and tumbles in most uncomfortable fashion, while the ship is
rarely if ever brought to a full stop, and the pilot, watching his
chance, must grasp the rope ladder let down its side, and scramble
aboard as best he can. Sometimes he gets a ducking, and if the
weather is tempestuous he is pretty certain to be drenched, but for
that he cares not a jot, and he is sure to show a smiling face to
captain and passengers when finally he sets foot on deck. Dropping
a pilot from an outgoing vessel is often more hazardous, especially
in stormy weather, than his transfer the other way. Then he must
descend the rope ladder and jump for the boat in the nick of time,
for to miscalculate in the least the position of the little shell means a
ducking almost certainly, and possibly a watery grave.
 PILOT SIGNALING A VESSEL
A peril, however, more feared by pilots than the one I have been
describing, is the dreaded lee shore; and with reason, as a story told
by a veteran ocean pathfinder will show. On a still afternoon in
midsummer the crew of a pilot-boat sighted a ship off Fire Island,
some five miles away. In the dead calm prevailing the only way to
board her was to row over the distance. There would be little danger
in doing this if the wind did not spring up and the ship sail away, so
the yawl was lowered and headed for the distant merchantman. But
as night was closing in, and ere the yawl had come within hailing
distance of the ship, of a sudden the breeze sprang up, and the
vessel making sail, glided slowly over the horizon line. The breeze
grew into a gale, and in the gathering storm and gloom the men
could no longer discern the whereabouts of the pilot-boat. Nor, there
being no compass on board the yawl, could they determine the
direction in which they were being blown. The nearest land was
miles away and the only thing that could be done was to keep the
boat's head to the wind and wait. Thus the minutes lengthened into
hours. Toward dawn, when the night was darkest, they heard the
thunder of surf on the reefs, and a little later felt the yawl lifted up
on the crest of a mighty breaker rushing swiftly toward the land.
There was a deafening roar, a crash, a whirl, and a torrent of foam.
In a twinkling the boat was capsized and the poor fellows were
struggling in the surf. One struck a rock and was killed. The others,
freed from the receding wave, ran up the beach, and by digging
their hands into the sand to escape the deadly undertow, finally got
ashore, drenched and exhausted.
In the main, however, the system I have been describing has now
become a thing of the past. Potent causes have contributed to this
result. Formerly pilot-boats had no particular stations assigned to
them, and boats have been known to cruise as far north as Sable
Island, a distance of six hundred miles, in order to get steamers
taking the northern courses. In the same way pilot-boats cruised
long distances to the southward and straight out to sea to meet the
incoming steamers and sailing vessels. Thus, unrestrained in its
movements and left to seek out its own salvation, each boat sought
to outdo the other in securing work, and all sorts of strategic devices
were brought into play in order to first gain the side of an incoming
vessel. Pilots took advantage of fog and night in order to slip by a
rival, while jockeying for winds and position was indulged in to an
extent that would be counted extraordinary in a yacht race.
Competition, however, cut down earnings to such an extent that
there came a time when many of the boats were no longer able to
pay expenses. Then it was that some of the long-headed among the
pilots, casting about for a remedy for this evil, came to the
conclusion that one steam pilot-boat would be able to do the work of
three or four sailboats. It was accordingly decided some years ago
that steamboats should gradually replace the existing fleet of sail.
With this innovation came restrictions regulating the cruising
grounds of the boats. Instead of cruising about indiscriminately as
formerly, each boat is now assigned a certain beat. An imaginary arc
has been described extending from Barnegat to Fire Island, a
distance of seventy-five miles, and all pilot boats are expected to
confine themselves within this line. Four pilot-boats patrol this line,
each covering a beat of about nineteen miles. Inside of the circle are
stationed two more pilot-boats, while still further in is a boat known
as the inner pilot-boat. Just off the bar another boat is stationed to
receive the pilots dropped by outward-bound vessels. When a boat
in the outer circle becomes unmanned or disabled, a boat from the
inner circle takes its place, while a reserve boat occupies the beat
left vacant on the inner circle. In this way all the beats are
constantly patrolled in an efficient and economical way. Each pilot
takes his turn at the service, and is on board a boat cruising on the
stations three days in seven, a moving contrast to the offshore
service of other years, when a boat and crew were frequently
compelled to remain at sea for weeks at a time.
Indeed, under the new system of pilotage, battles with cross-seas
and gales and exposures to snow, cold and sleet, while cruising for
vessels hundreds of miles off coast, are fast becoming things of the
past, and for stories of collisions, wrecks, narrow escapes and
strange mishaps, one must now hark back to the records of former
days. Here, however, he is sure to encounter many a tale that
quickens the pulse and stirs the blood. Take the case of the
Columbia, run down by the steamship Alaska, off Fire Island. When
the Alaska was sighted, the pilot-boat was head-reaching to the
north on the port tack. The wind was blowing a gale from the
northwest, and an ugly sea was running, with the weather clear, but
cold. She plunged deeply into the heavy sea, and heeled to the force
of the wind until her lee rail was awash. The wind whipped off the
top of the waves and filled the air with spray. When the steamship
sighted the boat off Fire Island, her course was changed to make a
lee for the boat's yawl. She seemed to stop when the yawl was
launched and two men and a pilot went over the side of the boat
and dropped into her, but ere the yawl had fairly started on her way
the liner, of a sudden, and without warning, forged ahead. The surge
from the port bow of the Alaska, as she pitched into a big wave,
capsized the boat, and threw the men into the water. Before
anything could be done to save them the bows of the steamship
rose and fell again, and, hitting the pilot-boat, cut it in two and
crushed the decks and beams to bits, the broken timbers being
swept under the bows and along the sides as the steamship again
forged ahead and passed over the spot. Not a man on the Columbia
was saved.
The Sandy Hook pilot, however, never quails in the face of danger or
even death, as was proved at the stranding of the packet boat, John
Minturn, almost within a stone's throw of the New Jersey beach
during a frightful hurricane in February, 1846. There were fifty-one
souls on board the Minturn, and of that number only thirteen
escaped to tell the story of that fearful night. Its hero, according to
the evidences of all, was Pilot Thomas Freeborn, who to the very last
struggled manfully to succor the hapless women and children who
clung to the deck around him. It was bitter cold, and every wave
that washed over the stranded ship left its coating of ice on deck,
rigging, passengers and crew. Freeborn and brave Captain Stark,
who was forced to see his wife and children freeze to death without
being able to render them assistance, gave up their own clothing in
a vain attempt to protect the weaker sufferers, and when days
afterward the pilot's body was found washed up on the beach it was
almost naked, while that of a woman, which lay near-by, was
carefully wrapped in his pea-jacket.
It has been three-score years since the wreck of the Minturn, but in
every year since then there has been numbered among the
members of the Sandy Hook Pilot's Association scores of hardy men,
who, should need come to them, stood ready to risk their lives and
die as bravely as did Thomas Freeborn. Pilot Henry Devere proved
that he had the same heroic fiber in his makeup when he sailed in
the James Funck, before the Civil War. A brig under shortened sails
was sighted one day, and when the yawl of the pilot-boat drew
alongside, Devere hailed a boy at the wheel. The boy seemed to be
stupefied, and the pilot was obliged to hail him several times before
he started up, leaned forward into the companionway, and called
feebly to somebody below. Then a gaunt man came on deck and
said that the crew had been stricken by fever. Most people in the
face of a menace of this sort would have turned back, but Devere
was not that kind of man. Instead, he went on board, and, with the
help of the mate, headed the vessel toward Sandy Hook. The
captain was ill in his stateroom. The body of a dead sailor found on
deck was tied in mosquito netting and dropped overboard. The boy
died in the lower bay, and the captain off the Battery, leaving the
mate as the sole survivor of the crew. The pilot helped to furl the
sails and make the lines fast, and only left the stricken vessel when
she had reached her moorings.
The stranding of the Jesse Carll in 1889, illustrates another of the
dangers with which pilots sometimes have to contend. The boat,
having discharged one of her five pilots, was standing off shore near
Fire Island, when she began to feel the force of an advancing
southern cyclone, and early in the evening was in what sailors call
"nasty weather." At midnight a violent thunder-storm burst
overhead, and the increasing wind raised a furious sea, but Pilot
Gideon Mapes, in charge of the vessel, had her under double-reefed
sails, and standing up against the wind and waves in fine shape.
Then came a deluge of rain, and the wind increased to hurricane
force. Soon a thick mist covered the water and shut out everything
in sight. The boat reached off and on, expecting to keep out of shoal
water, but all efforts failed. Her signals of distress were seen by the
life-saving crew on the beach, and before daylight the ten men on
board were taken ashore in boats. When morning came an effort
was made to pull the boat off, but as she shifted into deeper water
she filled, a hole having been made in her bottom. Then the pilots
abandoned her, but she was raised and repaired a few weeks later.
Stories like these are what the pilots tell in their idle hours.
Searching for them at such a time, one is most likely to find them at
the Pilots' Club, a flourishing social organization, which has roomy
quarters just under the roof of a big office building within hailing
distance of the Battery. Here at all hours of the day a score or more
of pilots are sure to be sitting about spinning yarns, playing cards
and checkers and reading the newspapers and magazines. Their
well-furnished clubrooms contain a great number of precious curios
—relics from all quarters of the globe. There are firearms of curious
antique pattern; autograph letters by such famous sea-dogs as
Macdonough and Porter; a tiny chest of drawers carved from one of
the timbers of John Paul Jones' ship, the Bon Homme Richard; a
portrait of Washington by Stuart, surrounded by two large American
flags, and a model of the pilot-boat Stingaree, which was built in
1810, and was one of the most famous crafts of her day.
This model shows that the years have wrought great changes in the
building and rigging of pilot-boats. In old times the boats simply
carried mainsail, foresail, and forestaysail and jib. They had no
foretopmast, and on their maintopmast carried a flying gaff-topsail,
which was hoisted from the deck. Now the boats have both fore and
maintopmasts, and each carries a mainsail, foresail, forestaysail, jib,
jib-topsail, maintopsail and staysail and fore and main standing-gaff
topsails, which give them an immense spread of sail, compared with
that used by the boats of earlier times. A schooner-rigged pilot-boat
costs from $15,000 to $16,000. That was about the cost of the
Caldwell H. Colt, a good example of the typical pilot-boat. She is
eighty-five feet long with twenty-one feet beam, 61.43 tons, custom-
house register, and a rig as trim and jaunty as that of an ordinary
yacht. The pride, however, at present writing, of the New York Sandy
Hook fleet is the New York, built of steel, propelled by steam, and
able to stand as much buffeting in cyclonic seas as the stanchest of
the liners. She was built on the Delaware from designs by A. Cary
Smith, is 155 feet long, 28 feet beam, 19 feet 7 inches deep, and is
driven by a compound surface-condensing engine of 100 horse-
power. Her pole masts are of steel, and she spreads on them enough
canvas to steady her. The New York has accommodations for twenty-
four pilots, who fare more luxuriously than they ever did on any of
the old sailing craft. They have a smoking-room in a separate steel
deckhouse, aft of the engine-room, fitted up like a similar room on
an ocean steamship, while the lifeboats in which they leave the New
York to board incoming vessels are hoisted and lowered by a steam
derrick in less than a minute. It is intended that in a few years the
entire fleet shall be made up of vessels equal if not superior to the
New York.
 CHAPTER VII
THE DEEP-SEA DIVER

There is something about the occupation of the diver that strongly

appeals to the imagination, and with reason, for working fathoms
below the surface of the water, in semi-darkness, dependent upon a
rickety pump for the breath of life, his trade is at best a perilous and
precarious one. Perhaps, that is why divers as a class are opposed to
taking apprentices, and that a majority of the men who drift into the
calling do so by accident. Most divers, if you question them, will tell
you that the best, if not the only way to acquire their art is to put on
a diving suit, go down into the depths, and learn the business for
yourself.
That was what a diver who was preparing for work in the East River
said to me, and, fitting the action to the word, I asked him to loan
me his suit, and permit me to try my 'prentice hand at the business.
He protested goodnaturedly, but finally yielding, brought out his suit,
and helped me to put it on. The outfit in which I speedily found
myself accoutred, consists of two suits, one within the other, and
both of india-rubber. The stockings, trousers and shirt are all made
together as one garment, which the wearer enters at the neck, feet
first. The hands are left bare, the wristbands of the rubber
shirtsleeves tightly compressing the wrists. There is a copper
breastplate, bearing upon its outer convex surface small screws
adjusted to holes in the neck of the shirt, which by means of nuts
fastened upon the screws, is held so securely in place as to render
the entire dress from the neck downward absolutely air and water-
tight. Fitting with equal closeness to the breastplate is a helmet,
completely inclosing the head and supplied with three glasses, one
in front and one on each side, to enable the diver to look in any
direction. Finally, for his feet there is a pair of very thick leather
shoes, made to lace up the front, and supplied with heavy leaden
soles to prevent him from turning feet uppermost in the water.
When, with my friend's aid, I had donned this curious-looking dress,
he placed across my shoulders ropes sustaining two leaden weights,
one hanging at my breast and the other at my back. Sometimes in
very strong currents it is necessary to make the weights which the
diver carries extraordinarily heavy. Such was the case with those
hanging over my shoulders on the occasion of my first dive. While
the diving dress I wore weighed of itself nearly two hundred pounds,
yet, much to my surprise, when once below the surface, I did not
find the burden I sustained in wearing it any more than I did that of
my ordinary clothing when out of the water. It also seemed
marvelous to me, after daylight had swiftly merged into the twilight
of the depths, that though I was several fathoms under water my
breathing was free and unconstrained, for an air-pump worked by
two men supplies the diver with air, which passes into his helmet
through a hose at the back. Near the place of its entrance is a spring
valve for its escape. This can be controlled by the diver, but he
usually sets it before going into the water and seldom disturbs it
afterward, since the pressure of the air being greater than that of
the water a surplus of the former readily escapes.
When the valve proves insufficient to permit the escape of all the
dead air the diver can open in his breast-plate a similar spring valve
intended only for such an emergency. He can also regulate the
amount of air pumped to him by signals on the air-hose to the men
engaged in pumping, one pull meaning more and two pulls less air.
These signals by means of the air-hose are generally used by all
divers, but each diver has also his own private code of signals upon
the life-line, which is always fastened to his waist, and by which he
is drawn up out of the water. These signals each diver writes down
very carefully and gives to the man in charge of the life-line. By
means of these he can, without coming to the surface, send for
tools, material or anything needed for the work he has in hand.
When a lengthy communication is to be made the diver often sends
up for a slate and writes what he wishes to say. Old divers declare
that it is just as easy to read and write under the water as it is out of
it, all objects being greatly magnified.
The only unpleasant sensation of my stay below was a slight
drumming in the ears—walking under the water I found an easy
matter—and when hauled to the surface I declared my first attempt
at diving a wholly successful one. However, the man whose suit I
had borrowed, smiled at my enthusiasm, and declared with
something akin to contempt that there was a good deal of difference
between deep-sea diving and grubbing about the East River for a
lost anchor. I learned before we parted that he was a deep-sea diver
forced for the moment to accept whatever task came to hand, but
there was truth in what he said; and I am also convinced, after talks
with a dozen members of his fraternity, that neither a single descent
nor even many descents into the depths, can give one an adequate
idea of the weird strangeness of a diver's life. That can come only
from the cumulative experience of a lifetime.
Almost all the submarine work on the Atlantic coast is done by divers
living in New York or Boston. There are about as many skilled divers
in Boston as New York—perhaps twenty in each city. The pay of a
skilled diver is five dollars a day of four hours or less. In that time a
man may descend half a dozen times, or he may descend once and
stay four hours, but be his period of labor long or short, it counts as
a day. If at the end of four hours he descends again that descent
counts as another day's labor. The diver's assistant receives three
dollars. He is a skilled man, whose business it is to manage the life-
line and the hose, and who sometimes becomes a diver. The
pumpers, who run the pump that keeps the diver supplied with air,
are each paid two dollars a day. They are not skilled workmen and
seldom develop into divers.
Probably a third of the New York divers do not work for wages.
These are men who own their outfits and prefer to work by the job.
Some of the self-employing divers enjoy good incomes from their
labors. As a rule, a diver of this class goes down, looks at a sunken
vessel, and then states what he will charge to raise her. Diver Victor
Hinston was paid $150 a day for locating the sunken steamship City
of Chester, and Captain Anthony Williams, having raised the
schooner Dauntless in two days, received $750 for his time and
trouble. The same diver, having repaired with iron plates and raised
in four days the steamer Meredith, ashore near Jeremie, in Hayti,
demanded and was paid $7,500 for his work. The divers of New York
live much as other citizens of the metropolis. A majority of them are
native Americans, with homes, wives and children. They are, of
course, absent from home a great deal and on short notice, for
divers from New York are not only sent all over the eastern coast of
the continent, but even to the Great Lakes and the interior rivers,
most of their work lying beyond the city.
Abram Onderdonk, when he died not long ago, was the oldest deep-
sea diver in this country. During forty of the nearly seventy years of
his life he was continuously engaged in the pursuit of his calling, and
it carried him to nearly every part of the globe. Captain Abe, as his
friends called him, counted the swordfish as the gravest danger
members of his craft have to fear. This fish, which has a short bony
sword almost as strong as steel, protruding from its head, speeds
along through the water, charging dead ahead and never veering
from its course for anything save a rocky ledge or the iron hull of a
steamship. If it strikes a wooden craft, its sword seldom fails to cut
clean through the vessel's side. Should a man be attacked by it
certain death awaits him. Diver Onderdonk himself never
encountered but one of these creatures, and that was a young one
whose sword had not yet hardened. He was at work on the deck of
a sunken vessel, when he saw the fish coming from a distance, and
heading straight toward him. He took a tighter grip upon the ax
which he held in his hand, and made ready for attack, but, to his
surprise and relief, the fish, never swerving from its course, glided
past him out of his guard's range, and a moment later disappeared.
Captain Abe often encountered sharks under water, but declared
that, as a rule, there is little to be feared from them. A former mate
of his named March, however, once had an ugly experience with
these creatures. The diver in question was at work in a wreck which
had been loaded with live cattle. When she had been at the bottom
for a month or so the cattle became light and began rising to the
surface. The locality was infested with sharks, which quickly
gathered round the hatchway, seizing the carcasses as they came
out and following them to the surface. Some of the cattle had been
tied, and these floating out to their ropes' end, were torn to pieces
by the sharks, which soon began to fight among themselves, with
the diver an unwilling witness to their struggles. March, hesitating to
ascend for fear he might be attacked, and afraid to remain below
lest the snap of a shark's mouth should sever his air hose, in the end
gave the signal to be hauled up, and the next instant was jerked into
and through the school of sharks. He came out of the water maimed
for life, as in his upward passage a shark snapped at him and took
off his right hand, thus rendering him incapable of further service as
a diver.
Another of Captain Abe's old mates, McGavern by name, while at
work in New Zealand waters, had an equally harrowing, although
fortunately less harmful, encounter with that most formidable of all
marine monsters, the devil fish. The diver was laying some wharf-
blocks when suddenly surprised by his uncanny foe. Despite his
struggles—and he was a giant in stature and strength—the monster
quickly and completely overpowered him. He was locked in the
tremendous claws of the devil fish, and fastened helpless against a
submerged spile. McGavern realized his peril, and kept quiet until his
assailant, whose arms measured nearly nine feet, loosened his hold.
Then he signalled to be drawn up, and came to the surface with the
writhing creature still clinging to his back.
Captain Abe served before the mast in his youth, and I find that,
other things being equal, sailors make the best divers of all. Their
former experience is apt to render them cool and quick-witted in the
presence of danger, and their knowledge of a ship's rigging and
construction proves of untold value to them in their work. To his
training as a sailor Captain Charles Smith, a well-known Boston diver,
probably owed his truly marvelous escape from death when
overtaken by accident while at work on the sunken hull of the Clara
Post, in the harbor of Bridgeport, Conn., a few years ago. The wreck
lay sixteen fathoms deep, and when Captain Smith descended to
examine it, he found that the masts had gone by the board, and that
the deck had been torn off by the waves, while the cross timbers
strewed with the wreckage, hung over the decks and into the hold.
Captain Smith began to cut them away, when suddenly the tangled
mass shifted and fell part way in the hold, catching him with it and
prisoning him as in a vise. The diver could not see far in the deep
water in which he was at work, and finding himself pinned in, how
he could not tell, he pulled the life-line three times—the signal that
his life was in peril. He felt himself rising a few feet; then all the
wreckage fell in upon him, pinning him more securely than before.
Worse still, when he tried to free himself, he found that the air-pipe
had encountered some unseen obstruction, and that to attempt to
move about would shut off his supply of air. The peril was one that
made each moment seem like eternity.
 A DIVER READY TO DESCEND
Meanwhile the diver's assistants were trying to discover what had
happened to him. It seemed to them that the signal to haul up had
been instantly followed by one to lower, and then by one to stop.
The men at the life-line, confused at these apparently contradictory
commands, ordered the derrick to haul on the blocks. Nothing
yielded to the strain, and the men at the pumps labored until they
were exhausted, and had to give way to others, but still no signs of
release. A new danger now threatened the imprisoned man. In
catching hold of some iron bolts he had cut a small hole in the valve
of one of his rubber gloves, and water, filling the glove, was slowly
oozing past the clamps at the wrist, and creeping up the arm. It
seemed to the helpless diver, held fast in the tide-swept mass, that
he would soon be strangled or crushed to death. Confused by the
great air pressure in his helmet, he had about concluded that his
end had come, when—unlooked for relief—the wreckage gave a
lurch, and he found that he could climb up to one of the deck
timbers. He grasped his ax, and was hewing desperately for
freedom, when suddenly the whole mass broke away, and began to
rise rapidly, carrying the diver, now head downward, with it. His
queer ascent did not consume more than ten seconds, but it was
long enough for him to live over in memory all the events of a
lifetime of two-score years. At first his comrades failed to discover
him in the mass of tangled material, and their surprise can be
imagined when he shot up through the wreckage, feet first. Captain
Smith described this as his closest call to death's door, "and" he
added, "I have peeped through the keyhole pretty often."
Captain Smith's adventure reminded a brother diver, in whose
presence it was told, of a narrow escape of his own. It occurred
while he was putting some copper on the bottom of a steamer in
dock. "I took some plate down with me," he said, "and worked for a
while on one side of the hull, after which I started in to put some
plates on the other side. The vessel was about three feet off the
bottom, and I crawled underneath, dragging the plates behind me.
After I had been at work for an hour or so I noticed that my air was
getting short, but when I tried to get under the keel again to be
hauled up, I found the steamer on the bottom and squeezing my air-
hose between its keel and the ground. The tide was ebbing and the
hull had gradually sunk until it was almost aground. I had forgotten
all about the tide, and when I pulled the hose it refused to move an
inch. If the bottom had been soft it would not have mattered so
much, but it was rock, and the hose was gripped like a vise. There
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

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

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com