Practical machine learning a new look at anomaly

detection First Edition Dunning - PDF Download

(2025)

https://ebookultra.com/download/practical-machine-learning-a-new-
look-at-anomaly-detection-first-edition-dunning/

Visit ebookultra.com today to download the complete set of

ebooks or textbooks
We have selected some products that you may be interested in
Click the link to download now or visit ebookultra.com
for more options!.

Depersonalization A New Look at a Neglected Syndrome 1st

Edition Mauricio Sierra

https://ebookultra.com/download/depersonalization-a-new-look-at-a-
neglected-syndrome-1st-edition-mauricio-sierra/

A First Course in Machine Learning Chapman Hall CRC

Machine Learning Pattern Recognition 2nd Edition Simon
Rogers
https://ebookultra.com/download/a-first-course-in-machine-learning-
chapman-hall-crc-machine-learning-pattern-recognition-2nd-edition-
simon-rogers/

A First Course in Machine Learning 1st Edition Simon

Rogers

https://ebookultra.com/download/a-first-course-in-machine-
learning-1st-edition-simon-rogers/

Innovative Minds A Look Inside Siemens Idea Machine German

Edition Ulrich Eberl

https://ebookultra.com/download/innovative-minds-a-look-inside-
siemens-idea-machine-german-edition-ulrich-eberl/
Data Mining Practical Machine Learning Tools and
Techniques 2nd Edition Ian H. Witten

https://ebookultra.com/download/data-mining-practical-machine-
learning-tools-and-techniques-2nd-edition-ian-h-witten/

Why Sex Matters A Darwinian Look at Human Behavior Low

https://ebookultra.com/download/why-sex-matters-a-darwinian-look-at-
human-behavior-low/

Global capitalism at bay 1st Edition Professor John H

Dunning

https://ebookultra.com/download/global-capitalism-at-bay-1st-edition-
professor-john-h-dunning/

Probabilistic Machine Learning for Civil Engineers James-A

Goulet

https://ebookultra.com/download/probabilistic-machine-learning-for-
civil-engineers-james-a-goulet/

Plato s immoralists and their attachment to justice A look

at Thrasymachus and Callicles First Edition Peter Jerrold
Hansen
https://ebookultra.com/download/plato-s-immoralists-and-their-
attachment-to-justice-a-look-at-thrasymachus-and-callicles-first-
edition-peter-jerrold-hansen/
Practical machine learning a new look at anomaly
detection First Edition Dunning Digital Instant Download
Author(s): Dunning, Ted;Friedman, B. Ellen
ISBN(s): 9781491914182, 1491914181
Edition: First edition
File Details: PDF, 10.40 MB
Year: 2014
Language: english
Practical Machine Learning
A New Look at Anomaly Detection

Ted Dunning and Ellen Friedman

Practical Machine Learning
by Ted Dunning and Ellen Friedman
Copyright © 2014 Ellen Friedman and Ted Dunning. All rights reserved.
Printed in the United States of America.
Published by O’Reilly Media, Inc., 1005 Gravenstein Highway North, Sebastopol, CA
95472.
O’Reilly books may be purchased for educational, business, or sales promotional use.
Online editions are also available for most titles (http://safaribooksonline.com). For
more information, contact our corporate/institutional sales department: 800-998-9938
or [email protected].
Editor: Mike Loukides

June 2014: First Edition

Revision History for the First Edition:

2014-05-14: First release

2014-08-08: Second release

See http://oreilly.com/catalog/errata.csp?isbn=9781491911600 for release details.

Nutshell Handbook, the Nutshell Handbook logo, and the O’Reilly logo are registered
trademarks of O’Reilly Media, Inc. Practical Machine Learning: A New Look at Anomaly
Detection and related trade dress are trademarks of O’Reilly Media, Inc.
Many of the designations used by manufacturers and sellers to distinguish their prod‐
ucts are claimed as trademarks. Where those designations appear in this book, and
O’Reilly Media, Inc. was aware of a trademark claim, the designations have been printed
in caps or initial caps.
Photos are copyright Ellen Friedman.
While every precaution has been taken in the preparation of this book, the publisher
and authors assume no responsibility for errors or omissions, or for damages resulting
from the use of the information contained herein.

ISBN: 978-1-491-91160-0
[LSI]
 Table of Contents

1. Looking Toward the Future. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2. The Shape of Anomaly Detection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Finding “Normal” 8
If you enjoy math, read this description of a probabilistic
model of “normal”… 10
Human Insight Helps 11
Finding Anomalies 12
Once again, if you like math, this description of anomalies
is for you… 13
Take-Home Lesson: Key Steps in Anomaly Detection 14
A Simple Approach: Threshold Models 14

3. Using t-Digest for Threshold Automation. . . . . . . . . . . . . . . . . . . . . . 15

The Philosophy Behind Setting the Threshold 17
Using t-Digest for Accurate Calculation of Extreme
Quantiles 19
Issues with Simple Thresholds 20

4. More Complex, Adaptive Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Windows and Clusters 25
Matches with the Windowed Reconstruction: Normal
Function 28
Mismatches with the Windowed Reconstruction:
Anomalous Function 30
A Powerful But Simple Technique 32
Looking Toward Modeling More Problematic Inputs 34

iii
5. Anomalies in Sporadic Events. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Counts Don’t Work Well 36
Arrival Times Are the Key 38
And Now with the Math… 40
Event Rate in a Worked Example: Website Traffic Prediction 41
Extreme Seasonality Effects 43

6. No Phishing Allowed!. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
The Phishing Attack 47
The No-Phishing-Allowed Anomaly Detector 49
How the Model Works 50
Putting It All Together 51

7. Anomaly Detection for the Future. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

A. Additional Resources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

iv | Table of Contents
 CHAPTER 1
Looking Toward the Future

Everyone loves a mystery, and at the heart of it, that’s what anomaly
detection is—spotting the unusual, catching the fraud, discovering the
strange activity. Anomaly detection has a wide range of useful appli‐
cations, from banking security to natural sciences to medicine to mar‐
keting. Anomaly detection carried out by a machine-learning program
is actually a form of artificial intelligence. With the ever-increasing
volume of data and the new types of data, such as sensor data from an
increasingly large variety of objects that needs to be considered, it’s no
surprise that there also is a growing interest in being able to handle
more decisions automatically via machine-learning applications. But
in the case of anomaly detection, at least some of the appeal is the
excitement of the chase itself.

1
Figure 1-1. Finding anomalies is the detective work of machine
learning.

When are anomaly-detection methods a good choice? Unlike fictional

detective stories, in anomaly detection, you may not have a clear sus‐
pect to search for, and you may not even know what the “crime” is. In
fact, one way to think about when to turn to anomaly detection is this:
Anomaly detection is about finding what you don’t know to look for.
You are searching for anomalies, but you don’t know what their char‐
acteristics will be. If you did, you could use a different form of machine
learning, called classification, or you would just write specific rules to
find the anomalies. But that’s not generally where you start.
Classification is a form of supervised learning where you have exam‐
ples of each kind of thing you are looking for. You apply a learning
algorithm to these examples to build a model that can use features of
new data to classify them into categories that represent each kind of
data of interest. When you have examples of normal and some number
of abnormal situations, classifers can help you mark new situations as
normal or abnormal. Even when you know about some kinds of
anomalies, it is always good to keep an eye out for new kinds that you
don’t know about. That is where anomaly detection is applied.

2 | Chapter 1: Looking Toward the Future

So you use the unsupervised-learning approach of anomaly detection
when you don’t know exactly what you are looking for. Anomaly de‐
tection is a discovery process to help you figure out what is going on
and what you need to look for. The anomaly-detection program must
discover interesting patterns or connections in the data itself, and the
detector does this by first identifying the most important aspect of
anomaly detection: finding what is normal. Once your model does that,
your machine-learning program can then spot outliers, in other
words, data that falls outside of what is normal.
Anomalies are defined not by their own characteristics, but in contrast
to what is normal. You may not know what the anomalies will look
like, but you can build a system to detect them in contrast to what
you’ve discovered and defined as being a normal pattern. Note that
normal in this context includes all of the anomalies that you already
know about and have accounted for using a classifier. The outliers are
only those events that don’t match what you already know. Consider
this way to think about the problem: anomaly in this context just
means different than expected—it does not refer to desirable or un‐
desirable. You may know of certain types of events that are somewhat
unusual and require attention, perhaps certain failures in a system. If
these occur sufficiently often to be well characterized, you can use a
classifier to catalog them as problems of a particular type. That’s a
somewhat different goal than true anomaly detection where you are
looking for events that are rare relative to what is expected and that
often are surprising, or at least undefined ahead of time.
Together, anomaly detection and classification make for a useful pair
when it comes to finding a solution to real-world problems. Anomaly
detection is used first—in a discovery phase—to help you figure out
what is going on and what you need to look for. You could use the
anomaly-detection model to spot outliers, then set up an efficient
classification model to assign new examples to the categories you’ve
already identified. You then update the anomaly detector to consider
these new examples as normal and repeat the process. This idea is
shown in Figure 1-2 as one way to use anomaly detection.

Looking Toward the Future | 3

Figure 1-2. Use anomaly detection when you don’t know what to look
for. Sometimes this discovery process makes a useful preliminary stage
to define the categories of interest for a classifier.

Anomaly detection, like classification, is not new, but recently there

has been an increased interest in using it. Fortunately, there also are
new approaches to carrying it out effectively in practical settings;
much more accurate and sophisticated methods are now available.
Some of the biggest changes have to do with being able to handle
anomaly detection at huge scale, in real time. We will describe some
approaches that can help, especially when using a realtime distributed
file system. We will focus particularly on approaches that have demon‐
strated, practical, and simple implementations.
The move from specialized academic research to methods that are
useful for practical machine learning is happening in response to more
than just an increase in the volume of available data—there is also a
great increase in new types of data. For example, many new forms of
sensors are being deployed. Smart meters monitor energy usage in
businesses and residential settings, reporting back every few minutes.
This information can be used individually or looked at as a group from
a particular geographical location.

4 | Chapter 1: Looking Toward the Future

Figure 1-3. This wall of smart meters reports a granular view of energy
usage for a utility company. Sensor data is becoming a huge source of
valuable information that can be analyzed through machine learning
techniques such as anomaly detection.

Industrial equipment such as drilling rigs and manufacturing tools use

sensors to report on a wide range of parameters. The advances in
medical device sensors are astounding. Radio-frequency identifica‐
tion (RFID) tags are also commonplace on merchandise in retail
stores, in warehouses, or even on your cat. Data provided by these
sensors and other sources range from simple identification signals to
complex measurements of temperature, pressure, vibrations, and
more.
How can reporting from all these interconnected objects be used?
Collectively, these objects begin to make up the Internet of Things
(IoT). Relationships between objects and people, between objects and
other objects, conditions in the present, and histories of their condi‐
tion over time can be monitored and stored for future analysis, but
doing so is quite a challenge. However, the rewards are also potentially
enormous. That’s where machine learning and anomaly detection can
provide a huge benefit.

Looking Toward the Future | 5

Analysts predict that the number of interconnected devices in the In‐
ternet of Things will reach the tens of billions less than a decade from
this writing. Machine-learning techniques will be critical to our un‐
derstanding of what the signals from devices are telling us.
As we collect and analyze more data from sensors, we achieve a more
granular view of how our systems are functioning, which in turn gives
us the opportunity for a greater awareness of when things change for
better or for worse. Not only is there a growing need for more accurate
anomaly detection, there is also a growing desire for new and more
efficient ways to “cut to the chase” in order to be able to put anomaly
detection to work in practical, real-world settings. Practical anomaly
detection is more than just selecting the right algorithm and having
the technical expertise to build the system—it also means finding sol‐
utions that take into account realistic limitations on resources, sched‐
uling demands including time-to-value to make the projects cost ef‐
fective, and correct understanding of business goals.
In this publication, we show you the underlying ideas of why anomaly
detection works and what it’s good for. We explore the idea of finding
what is normal, deciding how to measure things that are far from nor‐
mal and how far that must be to be considered an outlier (Chapters 2
and 3). We provide a new method to do this (t-digest) and look at how
it can be applied in very simple systems (Chapter 3) and also in more
complex systems (Chapters 4 and 5).
Throughout this report, we strongly recommend the use of adaptive,
probabilistic models for predicting what is normal and how to contrast
that to what is observed. One of our topics in Chapter 4 dabbles in
deep learning with a time-series example, or at least dips its toe into
the shallow end of that pool. Although this is an advanced concept,
the execution of it in our example is surprisingly simple—no advanced
math required.
Chapter 5 provides some very practical ways to model a system with
sporadic events, such as website traffic or e-commerce purchases. In
Chapter 6, we provide a practical illustration of many of the basic
concepts in the form of detecting a phishing attack on a secure website.
Let’s see how all this works.

6 | Chapter 1: Looking Toward the Future

 CHAPTER 2
The Shape of Anomaly Detection

The exciting thing about anomaly detection is the sense of discovery.

You need a program that can spot what is unusual, so anomaly-
detection models are on the lookout for the outliers. To get a sense of
how this works, try a simple human-scale example, such as the one
shown in Figure 2-1. Can you spot an outlier?

Figure 2-1. Can you spot an anomaly in this data?

Despite the fact that there is apparent noise in the data of the horizontal
line shown in Figure 2-1, when you see data like this, it’s fairly easy to
see that the large spike appears to be an outlier. But is it?

7
What happens when you have a larger sample of data? Now your per‐
ception changes. What had appeared to be an anomaly turns out to be
part of a regular and even familiar pattern: in this case, the regular
frequency of a normally beating heart, recorded using an EKG, as
shown in Figure 2-2.

Figure 2-2. Normal heartbeat pattern recorded in an EKG. The spikes

that had, in isolation, appeared to be anomalies relative to the hori‐
zontal curve are actually a regular and expected part of this normal
pattern.

There’s an important lesson here, even in this simple small-scale

example:
Before you can spot an anomaly, you first have to figure out what “nor‐
mal” is.
Discovering “ normal” is a little more complicated than it sounds,
especially in a complex system. Often, to do this, you need a machine-
learning model. To do this accurately, you also need a large enough
sampling of data to get an accurate representation. Then you must find
a way to analyze the data and mathematically define what forms a
regular pattern in your training data.

Finding “Normal”
Let’s think for a moment about the basic ideas that underlie anomaly
detection, including the idea of discovering what is to be considered
a normal pattern of behavior. One basic but powerful way to do this
is to build a probabilistic model, an idea that we progressively develop

8 | Chapter 2: The Shape of Anomaly Detection

here and in Chapters 3 through 6. A good way to think about this is
in terms of mathematic symbols, but in case that’s not your preference,
consider the key ideas through this thought experiment.
Suppose you are studying birds in a particular location, and you ob‐
serve, identify and count how many birds and of what species pass by
a particular observation point over the course of days. An entirely
made-up example of what these observations might look like is shown
in Table 2-1.
Table 2-1. Bird watching provides a simple thought experiment to show
how a probabilistic model works. Once a new species was observed, we
watched for it on subsequent days. The synthetic data of this simplified
example helps you think about how, based on the observations you’ve
made, you could build a model to predict several things about what
you expect to observe on Day x.
Species Day 1 Day 2 Day 3 Day 4 Day 5 Day x
A 33 17 21 31 18 ?
B 7 1 2 3 3 ?
C 3 3 2 1 1 ?
D 5 3 0 0 0 ?
E 13 13 8 7 9 ?
F 1 0 0 0 0 ?
G 1 0 0 0 0 ?
H 3 3 5 1 6 ?
I - 2 1 0 1 ?
J - 1 0 0 0 ?
K - - - 1 0 ?
L - - - - 1 ?
Next - - - - - ?

Some species occur in fairly high numbers each day, and these species
tend to be observed every day. Several species occur much less com‐
monly. New species are seen almost every day, at least this early in the
experiment. You can predict several things for a subsequent day (or
short series of days):

• How many total birds you expect to see

• How many species you expect to see
• How many birds of each species will fly by

Finding “Normal” | 9
 • How many new, previously unobserved, species will be seen

This prediction is nicely captured in the form of a probabilistic model.

It assumes that all things (species) have at least some likelihood to
occur, that some are more likely than others, and some are extremely
rare (or not yet observed), and so the estimate of their likelihood will
be a very small value. You can even predict how many new species you
might expect on a given day, even if you cannot predict which species
they will be. You can assign a probability for each event or type of event
and thus describe in probabilistic terms what you estimate to be “nor‐
mal.”

If you enjoy math, read this description of a

probabilistic model of “normal”…
For those of you who prefer a mathematical way to describe things,
read on. For the rest of you, just skip this description and go to the
next section.
Suppose that our best guess of the probability of some observation i
from the set of all possible observation is πi. The true underlying
probability is pi. Because our model is a probabilistic model, the values
πi are constrained by definition in the following way:

What this means is that if we make πi large for some i, then we have
to make it smaller for some other i. Moreover, since the πi have to all
be non-negative, smaller means closer to zero but never less than zero.
The very deep mathematical inference that we can draw here is that if
we make the average value of –log πi as small as possible, then we can
prove that the estimated probabilities, πi, will be as close as possible to
the true underlying probabilities, pi. In fact:

where the maximum is only achieved if πi = pi for all i.

A rare event is expected to have a small value for πi, and thus the value
for –log πi will all be a relatively large positive number. You might think

10 | Chapter 2: The Shape of Anomaly Detection

that creates a problem, making the average –log πi large when our best
estimate of probabilities would need for it to be small. But remember
that the average is computed by weighting things according to their
probability, so that each rare thing will only have a small contribution.
Depending on the details of your data and what they represent, there
are various ways to prepare data for use in such a model and a variety
of appropriate algorithms from which to choose. We will describe
several options in upcoming chapters.

Human Insight Helps

Discovering a normal pattern requires more than just a good machine-
learning model. Part of the process of discovering normal involves
human insight: you must interact with the modeling process to decide
what makes sense in your own situation. The example of a single
heartbeat shown in Figure 2-2 makes this point. In an abstract sense,
the spike is anomalous as compared to the rest of the data. In our
example, just collecting more data, such as what’s shown in a full EKG,
is more than enough to recognize that the spikes are a normal part of
the heart function, but before seeing more data, an expert who knew
it was an EKG being analyzed would also tell you that the spike was
not an anomaly.
Our bird-watching thought experiment can also illustrate this point.
Suppose you observe a brown pelican—is this to be expected? A do‐
main expert would tell you yes, if you live near a coastal region of North
America, or no if you live for instance in the inland state of North
Dakota. Similarly, cedar waxwings are not seen for most of the year in
California but sightings suddenly become relatively commonplace
during a short period of migration. Of course the historical data would
reveal these fluctuations, and this is just an analogy, but the point is
that human insight from someone with domain knowledge is a val‐
uable resource to put a probabilistic model into the proper context.
Continuing with our bird-watching thoughts, human experience
might also inform you that some sequential events tend to be related.
If you see one pelican fly by, it’s reasonable to expect several more
almost immediately because they often fly in a sort of squadron. But
you would not expect a couple of hundred pelicans in rapid succession.
This knowledge suggests to us not only that sequential measurements
may not be entirely independent—an important concept—but also

Human Insight Helps | 11

that models may require several levels of complexity to be really ac‐
curate predictions of real-world events.
So the first step in anomaly detection—using your model to discover
what is normal—also requires human insight both to structure the
model mathematically as well as to interpret what aspect of a pattern
is of interest and whether or not it represents a reasonable view of a
normal situation.
In building a machine-learning model for anomaly detection, you
have to identify the best choice of data, figure out how to put it into a
form acceptable to your algorithm and then acquire enough data for
training your model. In other words you will use data initially to let
the model discover patterns that you will then need to interpret in
order to determine the baseline or normal situation. This may require
a number of adjustments to the algorithm you use before you end up
with something that makes sense. The more you know about the sit‐
uation being investigated, the more easily and accurately you can de‐
cide when your model has achieved the first goal of anomaly detection
by finding what is normal.

Finding Anomalies
A second level of human insight is needed once you’ve established
what is normal and begin to look for what is anomalous. For our EKG
example, anomalous behavior is not the fact that there are spikes but
the observation that their frequency fluctuates during an episode of
abnormal heart behavior, as seen in the data displayed in Figure 2-3.

12 | Chapter 2: The Shape of Anomaly Detection

Figure 2-3. Anomalies in the frequency of the heartbeat show up as
unevenly spaced spikes such as those between approximately 1206 and
1210 seconds in this EKG.

Figures 2-2 and 2-3 just illustrate the point that the basis for anomaly
detection is to establish what is normal and then compare new events
to that pattern or model. This two-step process can be done in a variety
of ways from very simple models of fairly straightforward systems that
use an assigned threshold to send alerts for potential anomalies (as
described in Chapter 3) or more sophisticated models that are adaptive
and can deal with complex or shifting situations (as explained in
Chapters 4 through 6. In all of these cases, you are comparing observed
behavior to what has been defined as normal.
Back to our bird-watching example: when the observations show ei‐
ther a big decrease in the overall number of common birds or the
appearance of a large number of very rare birds, your model should
flag the change as being anomalous.

Once again, if you like math, this description of

anomalies is for you…
Thinking in terms of a probabilistic model, very rare, anomalous
events will be assigned a much lower probability value than normal
events during training. As a result, during an anomaly when we ob‐
serve these rare events, the anomaly score will be large precisely be‐
cause our estimate of their probability is very low.
Because an anomalous event has a lower probability value than those
usually observed, the anomaly score, which is the negative log of the

Finding Anomalies | 13
probability value, –log πi, will be larger, possibly much larger than
usual. In other words, the anomaly score will be farther from the ideal
maximum of zero. This increase suggests that our model’s estimation
of normal is less well matched to actual anomalous events—thus high‐
lighting the occurrence of outliers.

Take-Home Lesson: Key Steps in Anomaly Detection

The overall message here is broadly applicable to different types of
anomaly detection, regardless of the complexity of the system and the
choice of algorithms that are used. These steps form a general guide‐
line to goals when you are trying to build your own anomaly detector.
Ask yourself these questions:

• What is normal?
• What will you measure to identify things that are “far” from
normal?
• How far is “far”, if something is to be considered anomalous?

A Simple Approach: Threshold Models

You must experiment to determine at what sensitivity you want your
model to flag data as anomalous. If it is set too sensitively, random
noise will get flagged, and with huge amounts of data, it will be essen‐
tially impossible to find anything useful beyond all the noise.
Even if you’ve adjusted the sensitivity to a coarser resolution such that
your model is automatically flagging actual outliers, you still have a
choice to make about the level of detection that is useful to you. There
always are trade-offs between finding everything that is out of the or‐
dinary and getting alarms at a rate for which you can handle making
a response. These considerations, as well as a useful new way to set a
good threshold, are the topic of Chapter 3.

14 | Chapter 2: The Shape of Anomaly Detection

Other documents randomly have
different content
[Contents]
CHAPTER VIII
 ST. KITTS AND THE GORGEOUS ISLE
Despite all their interest, neither Statia nor Saba held aught that
linked them with the buccaneers,—indeed, I doubt if these
adventurers ever visited either,—and so, dipping our colors to old
Fort Orange in memory of the salute the ancient guns gave the
Andrew Doria, we bore on to Basseterre, the port and capital of St.
Kitts.

To one who has seen only the more northern islands, St. Kitts is a
revelation,—a fascinating sight,—and even after one has viewed the
more southerly isles with their overpowering grandeur of mile-high
mountains and wondrous forests, St. Kitts still holds its own, for it
possesses charms unlike those of any other of the Caribbees.

From the cloud-draped summit of Mount Misery—dark and sinister,

four thousand feet above the sea—to the beaches rimmed with
creaming foam, St. Kitts is a glorious mass of green,—green of a
thousand shades and tints, from that of ripening cane to that of the
deep, shadowy ravines of its [137]mountain forests. Upward from the
sandy beaches and rugged bluffs sweep the broad cane-fields,
undulating over hill and dale and reminding one so strongly of the
downs of Sussex that one no longer marvels that the homesick
English settlers, weary of the long and tedious voyage, gazed with
brimming eyes upon this smiling, sun-bright isle. Vividly, tenderly
green are the fields of young canes, golden or russet the others,
sienna-red the plowed acres between, but all are drenched with
tropic sunshine, and all, from a distance, seem as well tended and
as regularly laid out as a great garden.
And everywhere are the palms. As far as eye can see, the swaying
coco-palms line the shores above the tumbling surf. Against the sky
the plume-topped cabbage-palms show their sharp silhouettes above
the lesser trees upon the mountain sides. For miles along the
winding, perfect roads the towering royal palms form avenues of
great columnar trunks and drooping, feathery fronds. They cluster
above the lowly negro huts or shade the great plantation homes
without discrimination and with equal beauty, and they nod like giant
feather dusters above the roof-tops of the town.

Massive, majestic, and mountainous is the northern portion of St.

Kitts, and it takes no very vivid [138]imagination to see in towering
Mount Misery the likeness of St. Christopher bearing the infant Jesus
on his shoulder which caused Columbus to name the island after his
own patron saint. But to the south the mountains with their dense,
forest-clad slopes give way to hills covered with endless acres of
cane, until at Basseterre the island is almost flat, and only isolated
rounded Monkey Hill breaks the rolling, down-like land.

Basseterre is a fittingly pretty town for this lovely island, with its red
roofs and its pastel-tinted houses shaded by palms above the
wonderfully colored sea, on whose calm surface ride gaily painted
sloops and schooners and bevies of rowboats of every color of the
rainbow.

But it must be admitted that there is very little of interest here.

There is a fairly attractive public garden; flowering shrubs and trees
are everywhere; there are pretty embowered residences, and the
people are friendly and hospitable. But there is nothing distinctive
about the place: it might be any one of a score of dolce-far-niente
tropical towns, and it is by no means either prosperous or over-
clean. Time was when St. Kitts was a well-to-do island; its planters
lived like princes or feudal lords, fleets of ships rode to anchor in its
harbor, and thousands of toiling blacks planted and cultivated
[139]and garnered the golden canes which sent a steady flow of
molasses and sugar from the isle and brought an equally steady flow
of golden sovereigns back to the Kittefonians’ pockets. But the
omnipresent and lowly beet spelled St. Kitts’s doom, as it spelled the
doom of many another sugar-producing land, and though the island
is by no means poverty-stricken, and during the late war became
prosperous for a time, the golden days of the past will never return.

Efforts have been made to win back prosperity with sea-island

cotton, citrous fruits, and other tropical products; but it is a hard
matter indeed to wean a sugar-planter from canes; and even those
who have taken up the cultivation of other things have not been
over-successful.

As a winter resort, St. Kitts is delightful, for it boasts a good climate

and a healthful one; its scenery is magnificent; it possesses splendid
motor-roads that completely encircle the island; it offers excellent
fishing and hunting, plenty of outdoor sports, and an active volcano,
Mount Misery, with a wonderful climb through the virgin tropical
forests to its crater.

The island has, like its fellows, had a checkered career, but it can
boast of being the first of the British West Indies to be settled by the
English, [140]who established themselves here in 1623. However,
they did not succeed in holding it in undisputed possession, and
what with the Caribs, the pirates, and the French, those earlier
colonists had a mighty hard time of it. More than once St. Kitts came
wholly under the sway of France. At other times the two nations
buried the hatchet temporarily, and while the English confined
themselves to the northern half of the isle, with their headquarters
at Sandy Point, the French were content with the other half, with
Basseterre as their port; yet there was constant friction, and not
until 1782 was St. Kitts definitely turned over to the British.

Except in the name of the capital, there are few if any traces of
French occupancy, but at Brimstone Hill, close to Sandy Point, are
the massive ruins of extensive fortifications built by the British. Here,
on an isolated, precipitous mass of rock, for all the world like a
young mountain gone astray, is a solid mass of loopholed and
battlemented masonry completely covering every available portion of
the eight-hundred-foot hill. It is an impressive and redoubtable
fortification, well-nigh impregnable in the days of muzzle-loading
cannon and black powder, and complete with sally-ports, moats, and
drawbridges. But it is quite deserted and useless, the abode of
countless monkeys—descendants [141]of apes brought from Gibraltar
as pets, by the garrison—which are eagerly hunted and esteemed by
the Kittefonians as a great delicacy.

Looking upon this stupendous work of defense, one marvels that any
enemy ever dared attack or even approach St. Kitts, but, to tell the
truth, it never saw battle, for it was not built until 1793, ten years
after France and England ceased quarreling over the island and its
neighbors, and all too late to be of any value whatsoever to its
builders.

As an imposing ruin it is well worth a visit, but its historical

attractions are nil. Indeed, St. Kitts seems strangely lacking in
anything connecting its present with its rather turbulent and not at
all bloodless past. I do not think there is even the customary tale of
buried treasure on the island; at least I have never heard one. The
people do not claim to have found pirates’ hoards in their cane-fields
or caverns; and they do not even associate the name of any great
pirate chieftain with their delightful home.
Nevertheless, St. Kitts was at one time a resort of pirates—or, rather,
buccaneers—who were attacked by the Spaniards in 1629 and driven
from the island. Many of these, of French blood, made their way to
Tortuga, off the coast of Haiti, and there formed the nucleus of that
famous headquarters [142]of the Brethren of the Main. As far as
records go there is nothing to show that the freebooters ever
returned in large numbers or for an extended stay, and, according to
a most interesting document which I was so fortunate as to acquire
in St. Kitts, when they did appear they were met with so warm a
welcome that it is not at all surprising they gave the lovely island a
wide berth.

The faded and crumpled bit of worm-eaten parchment, which I value

even more than the ancient coins so opportunely acquired at
Anegada, is a fragment of the court records of St. Kitts in the good
old days, and its scarcely legible writing relates the following:

An assize and generall Gaole delivrie held at St. Christophers Colonie from
ye nineteenthe daye of Maye to ye 22n. daye off ye same Monthe 1701
Captaine Josias Pendringhame Magustrate &C. The Jurye of our
Soveraigne Lord the Kinge Doe presente Antonio Mendoza of Hispaniola
and a subjecte of ye Kinge of Spain for that ye said on or about ye 11
Daye of Apryl 1701 feloneousely delibyrately and malliciousley and
encontrarye to ye laws off Almightie God and our Soveraigne Lord the
Kinge did in his cuppes saucely and arrogantyly speak of the Governour
and our Lord the Kinge and bye force and armes into ye tavernne of John
Wilkes Esq. did entre and there did Horrible sware and cursse and did
felonoslye use theattenninge words and did strike and cutte most
murtherouslye severalle subjects of our Soveraigne Lord the Kinge. Of w’h
Indictment he pleadeth not Guiltie butte onne presente Master Samuel
Dunscombe mariner did [143]sware that said Antonio Mendoza was of his
knowenge a Bloodthirste piratte and Guiltie of diabolicalle practises & ye
Grande Inquest findinge yt a trewe bill to be tryd by God and ye Countrye
w’h beinge a Jurie of 12 men sworne finde him Guiltie & for the same he
be adjuged to be carryd to ye Fort Prison to haave both his earres cutt
close by his head and be burnet throughe ye tongee with an Hot iron and
 to be caste chained in ye Dungon to awaitte ye plesyure of God and Our
Soveraigne Lord the Kinge.

We cannot but pity the luckless Spaniard who under the spell of
Kittefonian rum, or possibly palm toddy, did “Horrible sware and
cursse” and who may very likely have been quite innocent of any
piratical or “diabolicalle” past, for the British had no love for the
Dons and even when nominally at peace with Spain thought little of
putting an end to any subject of the Spanish king who came their
way. No doubt the very fact that the prisoner was a Spaniard was his
undoing; and that worthy mariner Samuel Dunscombe probably
perjured himself for the satisfaction of seeing a Don tortured. At any
rate, it seems as though having both ears “cutt close” and having
one’s “tongee” perforated with a red-hot iron was pretty severe
punishment for the alleged crimes. But it only goes to prove how
times have changed, and how little we can judge, by present-day
standards, of what in those days was cruelty or inhumanity. [144]

Also in St. Kitts, though on another visit, I came into possession of

an equally interesting souvenir of olden times—a remarkable little
volume bearing the rather cumbersome title of: “The True Travels,
Adventures and Observations of Captain John Smith, in Europe, Asia,
Africa and America from Anno Domini 1593 to 1629; his accidents
and sea fights in the straits; his service and stratagems of war in
Hungaria, Transylvania, Wallachi, and Moldavia, against the Turks
and Tartars; his description of the Tartars, their strange manners and
customs of religions, diets, buildings, wars, feasts, ceremonies, and
living; how he slew the Bashaw of Malbritz in Cambria, and escaped
from the Turks and the Tartars; together with a continuation of his
General History of Virginia, Summer Isles, New England and their
proceedings since 1624 to this present 1629, published in Anno
Domini 1630.”
 ST. KITTS

Basseterre and Monkey Hill

 ST. KITTS

The Circus, Basseterre

 A DESCRIPTION of The South Sea & Coasts of AMERICA Conatining ye
whole Navigation and all those places at which Capt. Sharp and his
Companions were in the years 1680 & 1681

From all of which it will be gathered that our hero of Pocahontas

fame was an adventurer of many parts in divers lands, and that his
activities in Virginia were but minor incidents in his romantic career.
In fact, for a space, Smith was something of a pirate himself, judged
by our standards at least, and his accounts of sea battles and prizes
taken are fascinatingly quaint. Of St. Kitts, too, he has much to say,
and he gives us more of an insight [145]into the troubles and
tribulations of the first settlers on this “fayre islant” than any other
writer. Aside from raids by the man-eating Caribs, being harassed by
pirates, and constant quarrels with the French, the early English
settlers seem to have had a most unfortunate experience with
hurricanes, which Smith naïvely explains are “overgrowne and most
monstrous stormes.” Indeed, the very year of its settlement by the
English, 1623, a hurricane swept the island and wiped out the
settlers’ gardens, their tobacco-fields, their houses, and their fort.
Hardly had they recovered from this when, in September, 1625,
another hurricane hurled itself upon the island. This was even worse
than its predecessor of two years before, and Smith states that, in
addition to blowing down all the houses, the tobacco, and “two
drums into the air we know not wither,” it also “drove two ships on
shore, that were both split.” He adds: “All our provisions thus lost we
were very miserable, living only on what we could get in the wild
woods.… Thus we continued till near June that the Tortels came in,
1627.” Six months later the colonists were once more made
homeless through a hurricane, and until the end of the narrative
hurricane followed hurricane. 1 [146]

To-day, however, St. Kitts is by no means noted for its “overgrowne

stormes,” which may be of interest meteorologically as tending to
show that hurricanes are not so frequent in the Antilles as formerly
and may, in centuries to come, cease altogether.

Another interesting fact brought out by Smith is that St. Kitts was
largely populated by malefactors and convicts bought at so much a
head from British prisons, shipped to the West Indies like cattle, in
the stinking holds of small ships, and auctioned off as slaves among
the planters. When we stop to think of such things, of the
unspeakable atrocities practised by the masters or owners of these
unfortunates upon their own countrymen (whom they branded with
red-hot irons and mutilated or tortured on the least provocation), to
say nothing of compelling white men, women, and children to labor
half naked from sunrise to sunset in the cane- and tobacco-fields,
under a broiling sun and urged on by the cruel lash, the buccaneers
seem tender-hearted gentlemen by comparison.

South of this emerald isle, plainly visible from Basseterre and

separated from St. Kitts merely [147]by a narrow strait, lies Nevis. In
a gigantic, absolutely symmetrical cone of green the massive
volcano of Nevis rises against the sky, its brow crowned with a
perpetual diadem of drifting fleecy clouds and at its feet the
undulating green fields sloping to the sea.

Once the Mecca of the wealth and fashion of Europe and the
Antilles, the world’s most famous watering-place, a spot so thronged
with notables, so gay with great balls, state receptions, and palatial
gambling-resorts, so ablaze with silks, satins, and jewels, so flooded
with gold and riches that it became known as “The Gorgeous Isle,”
Nevis to-day is almost as dead as its volcano’s crater. And yet it is as
charming, its climate is as salubrious, its thermal springs and mineral
waters as life-giving, its fields and forests as alluring as in those days
when its harbor was thronged with stately ships and its streets and
hostelries rang to the song and laughter of satin-clad, bewigged
gentlemen, and ladies with powdered hair; and liveried negro link-
bearers lit the way for sedan-chairs ablaze with gilded scrolls and
cupids. But the vast estates, the palatial mansions, the great Bath-
House, and the marvelously appointed casinos are but memories—
crumbling ruins forlorn and overgrown. Nevis is but a ghost of the
“Gorgeous Isle” of the eighteenth [148]century, though a very
beautiful ghost.

Aside from its one-time fame as a spa, Nevis is mainly noted as the
birthplace of Alexander Hamilton and the place where Admiral Lord
Nelson was married. The house wherein our statesman was born still
stands on a hill near the town, though in a badly ruined state, and in
the ancient but well-preserved “Fig Tree Church” there is still the
thumbed and faded marriage register wherein one may read, under
the entries for the year 1787: “March 11, Horatio Nelson Esq.,
Captain of H. M. S. Boreas, to Frances Herbert Nisbet, widow.” What
a matter-of-fact record of the mighty, one-armed old sea-fighter’s
love romance!

But to my mind the most interesting thing in Nevis is the submerged

ancient capital of Jamestown, which in 1680 was destroyed by a
severe earthquake. Then, as though Nature wished to hide the ruin
she had wrought, the town with its tumble-down buildings, many of
its inhabitants, and—so it is said—vast wealth, sank bodily below the
sea. To-day, in calm weather, one may gaze downward through the
crystal-clear water and trace the faint outlines of coral-incrusted
walls of buildings that mark the resting-place of the drowned city.
[149]

History, unfortunately, has little information to give us concerning

Jamestown and its destruction, or of the events of that awful day.
One of the few survivors was a noted freebooter, a Captain Greaves,
—otherwise known as “Red Legs,”—who, having seen the error of
his ways, had abandoned his piratical career and had settled down in
Nevis to a life of peace in the guise of a well-to-do planter. But the
reformed pirate, being recognized and denounced by a former
victim, was arrested and cast into an underground dungeon, only to
be miraculously saved by the earthquake, which destroyed his prison
and heaved him up quite unharmed. Finding himself floating upon
the sea with the remains of the town fathoms deep beneath him, the
ex-pirate clung to a piece of wreckage, and after numerous
adventures safely reached another island, where he once more
essayed a respectable existence, and lived and died a highly
honored citizen.

This was but one incident in the romantic career of Greaves, or “Red
Legs,”—which is a far more appropriate name for a freebooter,—who
was a unique and fascinating character. Sold as a slave in Barbados,
as were thousands of Scotch and Irish prisoners taken in the days of
Cromwell, Greaves, in an effort to escape from a cruel master,
sought refuge on a Dutch ship in the harbor. By some
[150]mischance, he swam in the darkness to the wrong vessel and
found himself upon a pirate craft. Fate having thus taken a hand in
shaping his destiny, the erstwhile slave boy took to the buccaneers’
life as a duck takes to water. As all his unfortunate fellows were
known in the islands as “red legs,”—as their descendants are to-day,
—this new recruit of the pirates at once received the nickname,
which stuck to him through all his years of buccaneering. While he
was famous for his reckless daring, his almost uncanny luck in
piratical undertakings, yet he was never dreaded as were many of
his fellows. For Red Legs, despite his handicap, was a gallant and
chivalrous gentleman at heart, and though he scuttled ships and
sacked towns without end, yet he earned the reputation of never
harming women or putting prisoners to death or torture. He was, in
fact, that incredible paradox, a moral pirate, and in his declining
years he devoted large sums—whether honestly earned from his
plantation or loot from his piratical ventures is unknown—to charity
and churches.

The island of Nevis can boast of association with one other pirate,
who in a way was even more remarkable than Red Legs and
accomplished the most noteworthy feat in all the annals of
buccaneering. This was no less a personage than Bartholomew
[151]Sharp, who, after what was probably the greatest adventure
experienced by any of those most adventurous men the buccaneers,
sailed into Nevis, back in 1682, and, having decided to abandon the
sea and rest on his laurels, departed thence to England. [152]

This is an excerpt from a report by Captain Thomas Warner, [146]the founder of

1
the colony, who lies buried in Middle Island Church on the highway between
Sandy Point and Basseterre, where his tombstone informs us that he “boughte an
illustryous nayme with loss of noble blood.” ↑
[Contents]
CHAPTER IX
THE “DANGEROUS VOYAGE” AND THE EFFECT OF A
NAGGING TONGUE
No story of the West Indies in their relation to the buccaneers would
be complete without some mention of Bartholomew Sharp and his
marvelous cruise, which even Ringrose, his sailing-master and
historian, dubbed “The Dangerous Voyage.”

Of Sharp’s earlier days of pirating we know little, but that he was an

adept follower of the profession we may be sure, for his
contemporaries spoke of him as “that sea artist and valiant
commander” and, to use a slang expression, it took some pirate to
win such praise from the corsairs of the Caribbean.

At all events, Sharp evidently found the pickings of the Spanish Main
and its neighboring waters too poor for his liking, and, seeking richer
fields for his art, gathered together a wild and daring company of
some three hundred and fifty men and in April, 1680, sailed for the
Isthmus of Panama. Among this choice assortment of
companionable [153]spirits were many noteworthy pirates, for Sharp
had great deeds in view and aimed to outdo the redoubtable Sir
Henry Morgan himself. Ringrose, the historian of the buccaneers,
was there; Dampier the buccaneer naturalist 1; Wafer the surgeon;
Watling and Gayny; Jobson the chemist; Coxon and Sawkins and
many another. Reaching the isthmus, they disembarked and,
emulating Morgan, proceeded to cross the “Bridge of the World”
afoot by way of Darien, the wildest and hardest route. This in itself
was no mean task, but to the pirates it was only an incident, a
somewhat disagreeable means to an end and nothing more. Having
gained the shores of the Pacific, they promptly commandeered
canoes and without hesitation boldly attacked the Spanish fleet lying
in the lee of Perico Island, off the city of Panama.

Then followed a battle which must have satisfied even the most
bloodthirsty. As usual, the pirates won the day, captured all the
Dons’ ships, and, having thus secured the necessary tools of their
trade, they transferred armaments, ammunition, and such treasure
as there was, from the smaller ships to a four-hundred-ton galleon
known as La Santissima Trinidad or The Most Blessed Trinity. Finally,
having scuttled the craft they could not [154]use, they started on a
career of piracy which, as a record of successes, battles, murders,
mutinies, and bloodshed, has probably never been equaled.

Indeed, so execrable a pirate did Sharp prove himself that even

some of his most notorious fellows could no longer stomach him,
and Dampier, Gayny, Jobson, and over forty others deserted the
company and started back for the Caribbean via the Isthmus of
Darien.

Thereupon Sharp was seized with a brilliant idea, an inspiration

which had never come to his brother buccaneers, a wild scheme
quite worthy of his rash spirit. It was nothing less than to ravage the
entire western coast of South America, sail through the Strait of
Magellan, and return by sea to his old stamping-ground in the West
Indies.

This was “the dangerous voyage,” and while it proved far more
dangerous to the unfortunate peoples of the west coast than to the
buccaneers, yet the mishaps and adventures of the latter were
thrilling enough, and sufficiently numerous to fill a volume. No fiction
ever written, no imaginings no matter how vivid could equal
Ringrose’s log of The Most Blessed Trinity, as, sailing down the
coast, her crew landed and sacked towns, filled market-places with
dead and wounded, ravished women, pillaged cathedrals, razed
cities, and with [155]sword and torch left a trail of blood and
devastation from Panama to Patagonia.

Laden with loot,—with wines and spirits, silver bullion, 2 golden coins,
jewels torn from the fingers of terrified women; plate and chalices
from desecrated churches; embroidered vestments of murdered
priests, treasure won through unspeakable tortures; satins and silks;
even hides and tallow,—the battered galleon, scarred by shot and
shell, her gilded stern castle hacked away to afford room for guns,
her counter charred by fire, her decks blood-stained, cruised ever
southward toward the Horn.

But the buccaneers did not escape unscathed. At many a town they
were ignominiously defeated. Even Sharp had his troubles. As was
often the case with such rascals,—incredible as it may seem,—the
pirate crew, despite their ruthless and villainous lives, had certain
ideas of religion. Finding their captain utterly regardless of the
Sabbath, and so lacking in even a semblance of piety that he did not
hesitate to sack a town or scuttle a ship on that day, they decided
that the time had come to end such sacrilegious behavior, and,
seizing Sharp, they placed him in irons and dumped him into the
already [156]overcrowded hold. Then in his place they appointed a
new skipper, one John Watling, an hypocritical old villain who would
murder out of hand on Saturday and hold divine services the next
morning, when his cutthroat crew would join in singing hymns or
repeating prayers, the while wetting their throats with fiery rum. But
the consciences of the men were quieted, and the new captain
might have safely brought The Most Blessed Trinity safely to the
Antilles, had not Fate, in the shape of a bullet through his liver,
ended his sanctimonious and bloody career.

As there was no other capable of taking command, the mutineers

were compelled to reinstate Sharp, who apparently had been
meditating upon his own sins as he sat manacled in the dismal hold,
and with only his own thoughts for company had decided to lead a
better life henceforth. At any rate, one of the first things he did
when the irons were knocked from his wrists and ankles and he
found himself once more upon his ship’s quarter-deck, was to
intercede for the life of an aged Indian prisoner whom Watling,
before his sudden demise, had ordered shot for supposedly giving
false information about Arica.

Who would have imagined that the desperate and unconscionable

pirate chieftain,—whose greatest [157]enjoyment had been in the
screams of captured women, the shrieks of tortured men, the groans
of the dying and the roar of cannon—would ever come to this? But
the scarred, leather-skinned, shaggy-browed old villain waxed
eloquent and drew heartrending pictures of the poor Indian’s empty
home, of his wife and children seated in their lowly hut, waiting for
the return of their lord and master; and his chronicler even asserts
that the suddenly reformed Bartholomew’s voice faltered and tears
coursed down his cheeks as he spoke.

Unfortunately for the captive, the captain’s plea was in vain, and his
crew, still sullen and mutinous, being determined to put an end to
the old Indian, Sharp called for a basin of water and ostentatiously
washing his gnarled and blood-stained hands therein, wiped them on
the bedraggled and grease-spotted velvet coat he wore, and then,
with upturned eyes, declared most impressively and solemnly that
he was “clear of the blood of this poor man.” He added, “I will
warrant you a hot day for this piece of cruelty whenever we come to
fight at Arica,”—a prophecy which was fulfilled in a manner far
exceeding his expectations and believed by the freebooters, with
their sailors’ superstitions, to be the direct result of the Indian’s
death, as they later discovered he had told them nothing but the
[158]truth. At Arica scores of the invaders were killed or made
prisoners, among them the ships’ two surgeons, and a bare handful
of her original crew remained to work the badly battered and
strained ship, and reef and handle the patched and shot-riddled sails
as she staggered and plunged through the tempestuous, ice-filled
seas and freezing gales around the Cape and into the Atlantic
through uncharted waters. For they failed to make the Strait of
Magellan and won a way where no ship had sailed eastward before.
Possibly during his short confinement and his compulsory resignation
from leadership, Sharp really did find grace and decide to live an
honest life thenceforth; or more likely, being a canny rascal, he had
no desire to repeat his experience and determined to give his
rebellious crew their full of righteousness. Whatever the reason,
there was no further trouble, and The Most Blessed Trinity, having
successfully weathered the storms and billows of the Antarctic, went
wallowing on her way northward through the Atlantic.

Weather-beaten, storm-strained; her sails in tatters, her rigging gray,

ragged, and slack, her spars patched and fished in scores of places;
with yard-long weeds upon her leaking bottom, and bearing the
scars of many a battle, the one-time Spanish flag-ship worked up
the coast, through [159]the doldrums, and into the trades. Her log
reveals little but a constant succession of gales and hurricanes, of
starvation rations and ceaseless work to keep the old ship afloat and
able to sail, until, eighteen months after starting forth on her “most
dangerous voyage” she entered the Caribbean and sighted Barbados
—the first land seen since passing Patagonia.

But the sea-weary buccaneers were fated not to set foot upon the
right little, tight little isle. Within Carlisle Bay lay H. M. S. Richmond,
and Sharp, glimpsing the war-ship, promptly squared his battered
yards and headed for Antigua.
A dozen or more of the men, among them Ringrose, landed here—
eventually to make their way in safety to England—while The Most
Blessed Trinity, with men working ceaselessly at her wheezing
pumps, bore away for Nevis. Here, two years after Jamestown had
sunk beneath the waves, the wraith-like galleon with her lawless
crew came to rest, and the great “sea artist,” having accomplished
his harebrained undertaking, turned his battered old hulk over to his
fellows and, well laden with riches, sailed for England arrayed in all
the gorgeous finery of some murdered grandee.

Thus ended this most remarkable voyage, this greatest of buccaneer

adventures; a cruise unequaled [160]in the annals of the sea; the
longest, bloodiest, and most successful pirate raid of history.

No doubt, like Red Legs, Sharp settled down in some quiet nook and
spent the remainder of his life as a respected squire or gentleman
farmer in Sussex or Surrey; for nothing seems to be known of him
after he arrived in England, and, having been tried for and acquitted
of piracy, he dropped out of sight. But Basil Ringrose, to whom we
are indebted for the log of The Most Blessed Trinity, and who, as
navigator, was mainly responsible for the safe consummation of her
voyage, was far too restless a soul to be content with English lanes
and hedgerows and a vine-clad thatched cottage. Once more taking
to the buccaneer’s life, he joined a pirate ship bound for the South
Seas, and met death off the coast of Mexico after again rounding the
Horn.

It was not at all unusual for a pirate to give up the sea and, under
an assumed name, live quietly upon the fruits of his labors, with his
past quite unsuspected by his neighbors, but it was rarely indeed
that an honored and respected gentleman took suddenly to pirating.

Such, however, was the case with a certain Major Stede Bonnet, a
rich and finely educated citizen of Barbados and a pillar of the
church. [161]

He was a gentleman most highly thought of, a leader of Barbados

society, and apparently one of fortune’s favorites. But it seems that
there was a fly in the major’s ointment in the person of a nagging,
quarrelsome Mrs. Bonnet.

In fact, so unbearable did the major’s life become that at last he

decided to choose the less of two evils. Having purchased a sloop,
fitted her with guns, assembled a crew, and named his acquisition
The Revenge, Major Bonnet bade a thankful farewell to his wife and,
on a dark night in 1716, sailed forth from Bridgetown Harbor, bound
a-pirating.

Oddly enough, although the gallant if henpecked officer knew

nothing whatsoever of seamanship, he seems to have been a
somewhat successful and lucky pirate,—albeit a humane one, for it
was he who rescued the marooned members of Blackbeard’s crew
from their desert isle,—and The Revenge, cruising off our Atlantic
coast, took prizes right and left.

In a short time Bonnet’s name was one to conjure with, and a mere
mention of it brought terror to the hearts of shipping-men and
sailors in every port from Salem to Savannah, while so bold did he
become that he had the effrontery to make Gardiner’s Island, in
Long Island Sound, his headquarters [162]at various times. For a
space, too, he joined forces with the redoubtable Teach, better
known as Blackbeard, but the partnership was rather unfortunate,
for the scoundrelly Teach robbed the major of his ship and most of
his possessions.

This treacherous behavior on the part of a supposed friend made

Bonnet melancholy, if we are to believe his biographer, and may
have led to his undoing, for shortly afterward—in 1718, to be exact
—he was taken prisoner off the Carolinas. He managed to make
good his escape in a canoe, but a reward of seventy pounds sterling
was offered for him, and the following year he was captured at
Sullivan’s Island, was tried in Charleston, and, having been
sentenced to death, was hanged at White Point. Possibly poor Major
Bonnet was not sorry to find peace even at the end of the
hangman’s rope, for his adventures had brought him little more
comfort or ease than his home life in Barbados; and it is even
doubtful if his widow wept over his demise, or mended her ways. Of
her we know nothing; she slips quite out of the story with the
departure of her spouse on The Revenge, and even Johnson in his
“History of the Pyrates” was too gallant to do more than hint at her
character by stating that “This humour of going a pyrating was
believed to proceed from a disorder of the mind [163]said to have
been occasioned by some discomforts in the marriage state.”

However, we do know, from historical records, that poor Stede was

subjected to so lengthy a diatribe by the judge who condemned him
(it filled six closely written pages) that even his wife’s scoldings must
have seemed mild in comparison, and ere the judge’s long-winded
advice as to leading the higher life was ended the major must have
become so utterly worn out as actually to long for death.

Aside from Red Legs and the unfortunate Major Bonnet, Barbados,
or, for that matter, any of the southern islands, have little to link
them with the buccaneers; and as I was following in the wake of
these adventurers, and not endeavoring to make a cruise of all the
Antilles, the Vigilant sailed away from charming St. Kitts with Santo
Domingo as her goal.

There was a long sail ahead, nearly three hundred and fifty miles to
cover before we reached our objective point at Samana Bay, but we
planned to break the stretch by a stop at St. Croix, one hundred and
twenty-five miles from St. Kitts. Over a sparkling, sunny sea, with
the wind on our quarter, we sped away from the fair green hills and
downs of St. Kitts, left the great cone of Statia and the isolated,
beetling peak of Saba looming [164]hazily and like phantom isles to
the north, and with a broad wake of suds streaming far astern and a
roaring, curling mass of foam beneath our bows, swept across Saba
Bank.

It gives one a strange feeling to be sailing across the deep-blue

Caribbean far from land and suddenly to look over a ship’s rails and
plainly see bottom with its masses of corals, its great starfish, its
huge black sea-cucumbers, and its clustered sponges under the keel.
Many times have I sailed over Saba Bank, and never can I overcome
the impression that the ship is about to touch the ragged coral so
clearly visible; and the effect is even more remarkable when one is
crossing the bank on a big steamship. Saba Bank extends for nearly
forty miles east and west, and almost the same distance north and
south, with from six to twenty fathoms of glass-clear water over it;
and while in heavy weather large vessels avoid it, because of the
huge seas that pile up in the shoal water and the danger of touching
a reef, in calm weather they pass unhesitatingly across it.

Beyond Saba Bank we were out of sight of land, and on that wide
waste of waters stretching unbroken, even by a sail, to the horizon
on every side, the little Vigilant seemed pitifully small. More than
ever was I impressed with the courage, the [165]confidence, and the
faith in God which led the early voyagers and discoverers over this
unknown, uncharted sea in craft far smaller and less seaworthy than
the Vigilant. Rarely do we stop to consider what marvelous
undertakings and great adventures those ancient voyages were. In
miserable tubs—unwieldy, unfit for beating to windward with any
degree of success, clumsily sparred, and uncouthly rigged—the
daring navigators crossed the broad Atlantic and cruised hither and
yon among the reefs and isles of the Caribbean. The wonder is that
they did not all pile their timbers upon the ragged coral reefs and
low-lying cays, or that they ever managed to find the same island
twice. To steer a course across the Caribbean and raise one of the
isles above one’s bows is no small undertaking in a well-found vessel
and with the aid of modern instruments, and how Columbus and his
fellows ever accomplished it, with their crude appliances and with no
charts to guide them, is a mystery which ever fills me with wonder.

And as I lolled upon the Vigilant’s deck while our little ship with
flowing sheets plunged on to the westward, I also wondered how
big Sam could steer so unerringly for St. Croix—or if he could. But
he seemed to be perfectly confident of himself, [166]grasping the
wheel in those great black hands of his, casting an occasional glance
at the bellying sails, balancing himself on his huge sinewy legs to the
heave of the deck, and peering straight ahead with no heed to the
compass, as though his eyes could pierce the distance and see the
hills of St. Croix beyond the horizon.

Each day I took the sun and worked out our position, purely as a
precautionary measure; but I never revealed the results to Sam, for
I was curious to see just how accurately he could sail by dead
reckoning, or instinct, or whatever it was.

Once, as I stood near the taffrail and squinted through my sextant,

Sam chuckled, and a broad grin spread over his black, good-natured
face.

“Ah don’ guess you think Ah kin mek Santa Cruz, Chief,” he
ventured.

“Oh, I’m not worrying over that, Sam,” I replied. “But it’s a good
plan to know just where we are, in case of trouble—if a storm
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!

ebookultra.com