Introductory Statistics for Engineering

Experimentation 1st Edition Peter R. Nelson

download

https://ebookultra.com/download/introductory-statistics-for-
engineering-experimentation-1st-edition-peter-r-nelson/

Explore and download more ebooks or textbooks

at ebookultra.com
Here are some recommended products for you. Click the link to
download, or explore more at ebookultra.com

Using R For Introductory Statistics Second Edition John

Verzani

https://ebookultra.com/download/using-r-for-introductory-statistics-
second-edition-john-verzani/

Getting It Right R D Methods for Science and Engineering

1st Edition Peter Bock

https://ebookultra.com/download/getting-it-right-r-d-methods-for-
science-and-engineering-1st-edition-peter-bock/

Using R for Introductory Econometrics First Edition

Florian Heiss

https://ebookultra.com/download/using-r-for-introductory-econometrics-
first-edition-florian-heiss/

Using R for Introductory Econometrics 2nd Edition Florian

Heiss

https://ebookultra.com/download/using-r-for-introductory-
econometrics-2nd-edition-florian-heiss/
IBM SPSS for introductory statistics use and
interpretation 4th Edition George A Morgan

https://ebookultra.com/download/ibm-spss-for-introductory-statistics-
use-and-interpretation-4th-edition-george-a-morgan/

Introductory criminal justice statistics and data analysis

Second Edition. Edition Blevins

https://ebookultra.com/download/introductory-criminal-justice-
statistics-and-data-analysis-second-edition-edition-blevins/

Statistics with JMP Graphs Descriptive Statistics and

Probability 1st Edition Peter Goos

https://ebookultra.com/download/statistics-with-jmp-graphs-
descriptive-statistics-and-probability-1st-edition-peter-goos/

Statistics for Business and Economics Tenth Edition David

R. Anderson

https://ebookultra.com/download/statistics-for-business-and-economics-
tenth-edition-david-r-anderson/

Statistics for Business and Economics 11th Edition David

R. Anderson

https://ebookultra.com/download/statistics-for-business-and-
economics-11th-edition-david-r-anderson/
Introductory Statistics for Engineering Experimentation
1st Edition Peter R. Nelson Digital Instant Download
Author(s): Peter R. Nelson, Karen A.F. Copeland, Marie Coffin
ISBN(s): 9780125154239, 0125154232
Edition: 1
File Details: PDF, 17.50 MB
Year: 2003
Language: english
 Introductory Statistics for
Engineering Experimentation
by Peter R. Nelson, Marie Coffin, Copeland Karen A. F.

• ISBN: 0125154232
• Pub. Date: August 2003
• Publisher: Elsevier Science & Technology Books
 PREFACE

The Accreditation Board for Engineering and Technology (ABET) introduced a

criterion starting with their 1992-1993 site visits t h a t "Students must demonstrate
a knowledge of the application of statistics to engineering problems." Since most
engineering curricula are quite full with requirements in their own discipline, they
generally do not have the time for a traditional two semesters of probability and
statistics. A t t e m p t s to condense t h a t material into a single semester often result in
so much time being spent on probability t h a t the statistics useful for designing and
analyzing engineering/scientific experiments is never covered.
Therefore, we began to develop and teach a one-semester course whose purpose
was to introduce engineering/science students to some useful statistical methods.
We tried a number of different texts, but never found one we felt covered the right
material in a one-semester format. Our initial approach to covering the desired
material was to skip around in the text. That, however, was not well received
by many of the students and proved to be difficult for the instructor. In order to
alleviate some of the problems caused by skipping around, we started to supplement
the text with a set of class notes. This book is the outgrowth of those class notes.
We have used it as the text in our introductory statistics course for engineering and
science students at Clemson University since 1998. Continuous improvement since
then has resulted in the current book.
We have tried to use only as much theory as is necessary to motivate the tech-
niques. While calculus is a prerequisite for our course, it is actually only used when
obtaining probabilities, expected values, and variances from some simple density
functions. In line with the more recent A B E T requirement t h a t students must be
able to demonstrate "an ability to design and conduct experiments as well as ana-
lyze and interpret data" (Engineering Accreditation Commission, 1997) our main
emphasis is on t h e design and analysis of experiments. We introduce the idea of
modeling early (Chapter 3) by first modeling the deterministic aspects of an experi-
ment. Then (in Chapter 4) we model the random fluctuations (the random error)
around the deterministic part of the model. The remaining chapters put the two
modeling aspects together in order to do inference on the complete model, start-
ing with the simplest deterministic models and building up to the models most
useful in engineering experimentation. We emphasize the use of graphics in both

xi
xii PREFACE

the analysis and presentation of results. For example, normal probability plots are
used to assess normality. In this spirit we emphasize using the analysis of means
(ANOM) to analyze fixed-effect designed experiments when it is appropriate. In
our experience engineers and scientists find the ANOM much easier to understand
and apply than the analysis of variance (ANOVA). We also discuss the ANOVA
and point out that it is applicable to a broader range of problems than the ANOM.
All our students are required to do a project. Either individually, or in groups of
up to three, students must choose a topic of interest to them, design an experiment
to study that topic (we require at least two factors with one factor at more than
two levels), conduct the experiment, analyze the results, write a project report, and
give a presentation. The material is laid out so they are introduced to the design
aspects of experimentation early enough for them to design their experiments and
collect data for their projects while we are discussing exactly how this data should
be analyzed. Examples of some project reports based on actual projects are given
in one of the appendices.
We envision a course of this type being taught with some kind of statistical
software; however, the text is not oriented toward any particular software package.
All the data sets referred to in the text are available in a plain text format that can
be easily imported into virtually any statistics package. The examples and problems
reflect actual industrial applications of statistics.

Engineering Accreditation Commission (1997). "ABET Engineering Criteria 2000".

Accreditation Board for Engineering and Technology Inc., Baltimore, MD.
 Table of Contents

Preface

1 Introduction 1

2 Summarizing Data 6

3 Models for Experiment Outcomes 35

4 Models for the Random Error 84

5 Inference for a Single Population 155

6 Comparing Two Populations 217

7 One-Factor Multi-Sample Experiments 245

8 Experiments with Two Factors 295

9 Multi-Factor Experiments 338

10 Inference for Regression Models 369

11 Response Surface Methods 395

12 Appendices 424

13 References 508

Index 511
 1
INTRODUCTION

Most scientific disciplines are concerned with measuring items and collecting data.
From this book you will learn how to analyze d a t a in order to draw more powerful
conclusions from your experiments. W i t h the help of statistics, d a t a can be used
to:

9 Summarize a situation
9 Model experimental outcomes
9 Quantify uncertainty
9 Make decisions.

Because the quality of our conclusions depends on the quality of the original data,
some a t t e n t i o n will be given to efficient ways of collecting data. Generally, we
obtain d a t a by taking a s a m p l e from some larger population. T h e p o p u l a t i o n is
a conceptual group of all the items of interest to the scientist. We take samples
because the entire population m a y be infinite, or at least too large to be e x a m i n e d
in a timely and economic manner.

E x a m p l e 1.0.1 A production process makes seat posts for m o u n t a i n bikes t h a t

must fit into the seat post tube of a bike frame. T h e seat posts must be of a certain
diameter or they will not fit. T h e design specifications are 26 m m + 0.5 mm. To
monitor the production process a technician takes 10 posts off the p r o d u c t i o n line
every hour and measures them. If more t h a n 1 of the 10 are outside the specifi-
cation limits, the process is halted, and the controls t h a t determine d i a m e t e r are
investigated.

Frequently, questions of w h a t to measure and how to measure it are more dif-

ficult t h a n in the above example. W h e n the q u a n t i t y of interest is simple (like
diameter) it is obvious w h a t needs to be measured and how.
2 INTRODUCTION

E x a m p l e 1.0.2 Air pollution is a serious problem in m a n y cities and industrial

areas. In deciding what to measure one needs to consider the time, cost, and accu-
racy of different measurement systems. One good indicator of pollution is ozone
concentration. This can be measured fairly easily by taking an air sample and
counting the ozone in it (measured in parts per hundred million).

Variability
Clearly, if the answer to a statistical study is important, we should make our mea-
surements as precisely as possible. In fact, we often think of measured quantities
as completely "correct". No matter how much care is taken, however, the measure-
ment process will almost inevitably introduce some measurement error into our
calculations. That is, even if a careful person measures the same item repeatedly,
he will not arrive at the same answer every time. Furthermore, we are usually mea-
suring a new item each time, and the items (even if they appear to be identical)
will have their own inherent random variability. Because our calculations involve
b o t h of these two sources of variability, scientific conclusions always contain some
uncertainty. One of the functions of statistics is to quantify t h a t uncertainty. For
example, if we say a component will fail at an ambient t e m p e r a t u r e of 40~ it
matters if the estimate is 40 ~ + 1~ or 40 ~ + 10 ~

Experimental Design
In order to minimize the variability so t h a t it is possible to more precisely determine
what treatments in an experiment have significant effects, one must carefully plan
how the d a t a should be collected. Particularly when one is interested in studying
more than one factor at a time, such as studying the effects of different t e m p e r a t u r e s
and different times on a chemical reaction, much greater efficiency can be obtained
if the proper d a t a are collected.

E x a m p l e 1.0.3 Suppose one is interested in studying the effects of time and

t e m p e r a t u r e on the yield of a chemical process. The experimenter believes t h a t
whatever effect time has will be the same for the different temperatures. A c o m m o n
(but inefficient) approach is to study the effects of time and t e m p e r a t u r e separately
(this is referred to as a one-factor-at-a-time design). Such a design might call for
studying three times (for a fixed t e m p e r a t u r e ) and taking three observations for
each time. Then one might study two t e m p e r a t u r e s (for a fixed time) using three
observations for each temperature. W i t h these 15 total observations we can estimate
b o t h the average yield at a particular t e m p e r a t u r e and the average yield at a
particular time. The uncertainty in the average yield for a fixed t e m p e r a t u r e would
be less t h a n t h a t for a fixed time because more observations were taken (nine versus
six). (Exactly how to determine this uncertainty is discussed in Chapter 5.)
 INTRODUCTION 3

A much more efficient way to study the two factor (discussed in Chapter 8)
is to study them together. Taking two observations for each T i m e / T e m p e r a t u r e
combination would require only 12 total observations, but would provide estimates
for both the average yield at a particular t e m p e r a t u r e and the average yield at a
particular time with uncertainties of only 87% (= 100%V/9/12) of the uncertainty
for the average temperature from the previous design. Thus, by studying the two
factors together one can obtain better information with fewer observations.

Random Sampling
Even when studying only a single population, care must be taken as to how the d a t a
are collected. In order to quantify the uncertainty in an estimate of a population
characteristic (e.g., the diameter of a piston), one has to estimate the variability
associated with the particular characteristic. As already mentioned, this variability
consists of two parts: variability due to the characteristic itself (e.g., differences in
the pistons) and variability due to the measurement process. In order to correctly
assess the variability due to the characteristic of interest, one must collect samples
that are representative of the entire population. This is accomplished by taking
samples at random (referred to as random sampling). The idea is t h a t samples
should be chosen such that every possible sample is equally likely to be obtained.
While it is not always possible to perform exact random sampling, it is i m p o r t a n t
t h a t the sample be representative. For example, if one is interested in comparing
the lifetimes of two brands of batteries, one would want to obtain samples of the
batteries from different locations to help ensure t h a t they were not all manufactured
at the same time in the same plant. R a n d o m sampling ensures t h a t (on the average)
the samples are independent and representative. T h a t is, they are not related to
one another other than coming from the same population.

Randomization
In addition to random sampling the concept of randomization is also important.
The basic idea is t h a t treatments should be assigned to experimental units at ran-
dom and, when applicable, experimental trials should be performed in a r a n d o m
order so t h a t any biases due to the experimental units not behaving independently
because of some unknown or uncontrollable factor are on average distributed over
all the experimental trials.

E x a m p l e 1.0.4 Consider an experiment to compare rates of tire wear, and sup-

pose one is interested in only two brands. The tires are to be used on cars for say
10,000 miles of driving, and then the tread wear is to be measured and compared.
Since it is desirable to subject the two brands of tires to the same conditions affect-
ing wear, it would be reasonable to conduct a paired experiment (see the discussion
4 INTRODUCTION

on pairing in Chapter 6) and assign one of each brand of tire to each car used. Also,
since it is known that front and rear tires do not necessarily wear in the same way,
the two tires should both be mounted on either the front or the rear. Having chosen
say the rear, all that remains is to decide which brand goes on which side, and this
is where randomization comes into the picture. In order to average over b o t h brands
any effect on wear due to the side on which a tire is mounted, the brands should be
assigned to the sides at random. This could be done, for example, by using a table
of random numbers and associating odd numbers with the assignment of brand A
to the driver's side.

Replication
Multiple observations under the same experimental conditions are referred to as
replicates. Replicate observations are used to determine the variability under the
specific experimental conditions, and therefore, a true replicate must be obtained
from a complete repetition of the experiment. A replicate measure is not to be
confused with a repeated measure, where multiple measurements are taken on the
same experimental unit.

E x a m p l e 1.0.5 In Example 1.0.3 (p. 2) the two observations for each T i m e /

Temperature combination are two replicates. A true replicate in this case would
require the process being completely re-run under the same experimental condi-
tions. Simply running two assays on the product for a particular T i m e / T e m p e r a t u r e
combination would be a repeated measure rather than a true replicate since it would
only encompass the variability due to the assay (measurement error), and would
not reflect the variability in the process from batch to batch.
 INTRODUCTION 5

Problems
1. A scientist working with new fibers for a tennis ball covering wants to compare
the "bounce" of the balls with different coverings. One test will consist of
dropping balls with each covering from a specified height and counting the
number of times it bounces. Explain why it is better to use several different
balls with each covering, rather t h a n dropping the same ball repeatedly.
2. A student wants to measure the volume of liquid in cans of soft drink to see
how it compares to the nominal volume of 12 oz. printed on the can. Suggest
some sources of measurement error in this experiment, and discuss how the
measurement errors could be reduced or avoided.
3. An industrial engineer specializing in ergonomics performs an experiment to
see if the position (left, center, or right) within a work station affects the
speed achieved by a production line worker who packages a product from a
conveyor belt into boxes. Carefully explain what random error (i.e., r a n d o m
variability) signifies in this experiment. W h a t should be randomized in this
experiment and why?
4. Suppose you need to measure the size of some candies (Gummi Bears and
Peanut M&M's), where size is defined to be the m a x i m u m diameter of the
candies.
(a) Which kind of candy will probably be subject to more measurement
error? Explain.
(b) Suggest how this measurement error could be reduced.
(c) Which kind of candy will probably be subject to more r a n d o m vari-
ability? Explain.
5. Suppose a food scientist wanted to compare the "tastiness" of different vari-
ations of a cookie recipe.
(a) Suggest several ways to quantify (measure) "tastiness", and discuss the
advantages and disadvantages of each.
(b) Discuss the sources of variability (e.g., measurement error) in the con-
text of this problem, and suggest ways to minimize the effects of vari-
ability on the results of the study.
6. R&D chemists conducted an extensive study of a new paint formulation.
Four batches of the paint were made and ten test panels from each batch
were sprayed for testing. One characteristic of interest was the gloss of the
paint (an appearance measure). Discuss repeated measures versus replication
in the context of this study.
 2
S U M M A R I Z I N G DATA

In this chapter, we will consider several graphical and numerical ways to summarize
sets of data. A d a t a s u m m a r y should provide a quick overview of the important
features of the data. The goal is to highlight the important features and eliminate
the irrelevant details.
The first (and often most important) thing to do with a set of d a t a is to find
a way to display it graphically. Graphs can summarize d a t a sets, show typical and
atypical values, highlight relationships between variables, a n d / o r show how the d a t a
are spread out (what one would call the shape, or the distribution, of the data).
Even for small sets of data, important features may be more obvious from a graph
than from a list of numbers. We will discuss several ways to graph data, depending
on how the data were collected.

2.1 SIMPLE GRAPHICAL TECHNIQUES

Univariate Data
Histograms
W h e n d a t a are univariate (one characteristic is measured on each experimental
unit), a histogram is often a good choice of graph.

E x a m p l e 2.1.1 (Rail Car Data) A company ships many products via rail cars.
These rail cars are either owned or leased by the company, and keeping track of
their whereabouts is critical. The company is interested in minimizing the amount
of time t h a t a rail car is held by a customer. A histogram of the number of days
rail cars from a particular fleet were held by customers during a 4-month period
is given in Figure 2.1 and shows several interesting features of the d a t a (which are
found in r a i l c a r , t x t ) .
 SIMPLE GRAPHICAL TECHNIQUES 7

18

12
O
e'-
O
O-
O
U_

2 5 8 11 14 17 20 23 26 29 32 35 38 41 44
Days
Figure 2.1. Histogram of the Number of Days Rail Cars are Held.

9 T h e a m o u n t of time a customer keeps a rail car varies greatly. This makes schedul-
ing of the rail cars difficult.
9 There are a few observations in the right "tail" t h a t are far away from the major-
ity of the observations. Unusually low values are not of a concern in this situation,
however, unusually high values indicate a possible customer issue t h a t needs to
be resolved.
9 T h e number of days a rail car is held by a customer is centered a r o u n d 7 or 8.
9 Samples with long hold times are more spread out t h a n the samples with short
hold times. Thus, the distribution of hold times is skewed toward the higher
values, or right skewed.

Several steps are involved in constructing a histogram. You must decide on the
class boundaries, count the n u m b e r of observations in each class, and draw the
histogram. Generally, the class boundaries are chosen so t h a t somewhere between
10 and 20 classes are obtained. For larger d a t a sets, more classes are a p p r o p r i a t e
t h a n for smaller d a t a sets. There is no single correct way to construct the class
boundaries. All t h a t is required is an a p p r o p r i a t e n u m b e r of equal w i d t h classes
t h a t encompass all the data. One way to do this is to divide the range of the d a t a
(i.e., largest value - smallest value) by the desired n u m b e r of classes to o b t a i n an
a p p r o x i m a t e class width.

E x a m p l e 2.1.2 (Silica Surface-Area Data) Samples of a particular type of silica

(a chemical product with m a n y applications such as a filler in r u b b e r products) were
tested for their surface area (a key property). T h e resulting 32 m e a s u r e m e n t s are
8 SUMMARIZING DATA

listed in Table 2.1.

Table 2.1. Surface Areas for a Particular Type of Silica

101.8 100.5 100.8 102.8 103.8 102.5 102.3 96.9

100.0 99.2 100.0 101.5 98.5 101.5 100.0 98.5
100.0 96.9 100.7 101.6 101.3 98.7 101.0 101.2
102.3 103.1 100.5 101.2 101.7 103.1 101.5 104.6

The range of values is 1 0 4 . 6 - 96.9 - 7.7, and therefore to obtain approximately

10 classes, the class width should be 7.7/10 - 0.77 " 0.8. Starting with 96.9 as the
lower boundary for the first class, one obtains classes of 96.9-97.7, 97.7-98.5, and so
forth. The complete frequency distribution (a count of the number of observations
in each class) is given in Table 2.2. Note that since some of the d a t a values fall
exactly on a class boundary, one has to decide which class to put them in. It was
arbitrarily decided to put values that fell on the boundary in the upper class. For
example, the two 98.5 values were both put in the 98.5-99.3 class. The i m p o r t a n t
issue is not which class to put the value in, but to be consistent for all of the data.
The corresponding histogram is given in Figure 2.2.

Table 2.2. Frequency Distribution for the Data in Table 2.1

Class Frequency Relative Frequency

96.9-97.7 2 0.0625
97.7-98.5 0 0.0
98.5-99.3 4 0.125
99.3-100.1 4 0.125
100.1-100.9 4 0.125
100.9-101.7 8 0.25
101.7-102.5 4 0.125
102.5-103.3 4 0.125
103.3-104.1 1 0.03125
104.1-104.9 1 0.03125
Total 32 1.000
 SIMPLE GRAPHICAL TECHNIQUES 9

o
t--
(D

cr
CD
4
LI_

~
96.9 97.7 98.5 99.3 100.1 100.9 ~0~.7 ~oe.5 ~03.3 ~04.~ 104.9
S u r f a c e Area

Figure 2.2. Histogram for the Frequency Distribution in Table 2.2.

The histogram shows t h a t the surface area distribution is relatively balanced

about a middle value of around 101. Typically with d a t a such as this, m e a s u r e m e n t s
on multiple samples from the same source, one would expect more of a bell-shaped
distribution t h a n w h a t is seen here. However, 32 d a t a points is a small enough
sample t h a t the choice of class size or starting point can affect the look of the
histogram (see Problem 2.1.2, p. 15).

There are several things to note a b o u t histograms.

9 All of the classes have the same width.

9 T h e classes are adjoining, but not overlapping. This is so t h a t every observation
will be in one and only one class (i.e., it will get counted exactly once).
9 T h e n u m b e r of classes is s o m e w h a t arbitrary. Choosing too few classes will cause
your graph to lose detail, while choosing too m a n y m a y obscure the main features.
9 Histograms for small d a t a sets are more sensitive to the n u m b e r of classes t h a n
histograms for large d a t a sets.

Runs Charts
T i m e ordered plots or runs charts provide a view of the d a t a based on the order in
which the d a t a were collected or generated. Runs charts are of particular interest
when working with production data, however, they are not limited to such data. A
runs chart is simply a chart on which the d a t a are plotted in the order in which
they were collected. By plotting the d a t a with respect to time, one can check for
time dependencies.
10 SUMMARIZING DATA

E x a m p l e 2.1.3 (Historical Production Data) The runs charts in this example

were generated from historical data, that is, data that were collected and saved,
but analyzed after the fact; as opposed to data t h a t were generated in a controlled
experiment with the type of analysis to be performed on the data known in advance.
The runs charts were used to look at the process in the past to better understand
the process for the future.
Data from a chemical process were collected at different time intervals depending
on the particular variable being measured (some were collected every minute, some
every 5 minutes, etc.). The values associated with a particular variable were then
averaged over the shift. This resulted in one data point per shift and three data
points per day for each variable. Three plots are shown below, two for process
variables, and one for a quality variable. In reality this process had well over 100
process variables with recorded values and about a dozen quality variables. The runs
charts helped the plant engineers separate the variables that had high variability, or
were unusual, from all the others. For Process Variable 1 the runs chart (Figure 2.3)
shows a drift down over the first 30 time periods (i.e., 10 days), then the values
appear to stabilize. Such a drift is an event that the engineers might want to
research for causes, especially as it precedes the j u m p in the quality variable. The
runs chart for Process Variable 2 (Figure 2.4) does not exhibit any highly "non-
random" behavior, and thus, is of little interest. The chart for the Quality Variable
(Figure 2.5) shows a jump in the values at about time 35, but by time 70 it appears
to have been "corrected" (the target for this variable is 22).

13000

11000

9000

7000

20 30 4o so 60 10
Time

Figure 2.3. Runs Chart for Process Variable 1.

 SIMPLE GRAPHICAL TECHNIQUES 11

949.2

!
948.5

947.8

947.1
0 1'0 ~0 ~0 4o ~0 6o 7o
Time
Figure 2.4. Runs Chart for Process Variable 2.

24.1

23.3

22.5

21.7

20.9
0 10 20 30 40 50 60 70
Time
Figure 2.5. Runs Chart for the Quality Variable.
Random documents with unrelated
content Scribd suggests to you:
Sergeant, give Paitoto a rifle and detail a man to teach him to
shoot.”
Accordingly, on the 5th April, 1901, Captain Barton and I
marched out of Uiaku village in Collingwood Bay, in quest of the
Doriri, at the head of 159 men, 20 of whom were regular
constabulary, 6 village constables (armed), and about 50 Kaili Kaili
and Mokoru, the balance being composed of Maisina and
Collingwood Bay natives. I think that, up to this date, this was the
best organized and most carefully thought-out punitive expedition
that had ever been dispatched by a New Guinea Government. In one
respect, however, we were handicapped, and that was that, owing to
the non-arrival of the s.s. President with stores for the expedition, I
was obliged to purchase a quantity of rice from the miners (to whom
I have previously referred as being left in the Kuveri District, and
who were now abandoning their quest), and this rice, instead of
being packed in fifty-pound mats, was contained in sacks weighing
altogether seventy-five pounds, a cruel load for one man, and too
little for two carriers; unfortunately we had no extra mats or bags to
divide it up into again. The Kaili Kaili, however, came to my rescue,
by expressing themselves as able and willing to carry the heavy
bags, until they were reduced by daily consumption. The Kaili Kaili
and Mokoru were from first to last ideal carriers, never grumbling or
complaining at hard work, and quite prepared to follow anywhere or
do anything, and forming a pleasing contrast to the Maisina, who
began to suffer from nerves the moment that we had fairly set our
faces towards the country of the Doriri. We purposed sending back
the Maisina as soon as the food they carried was exhausted, and
then to rely entirely upon the Kaili Kaili and Mokoru.
The Maisina guided us by a winding and villainous track, across
a pestilential sago swamp, humming with mosquitoes; the track in
places was like a maze, for the purpose of confusing the Doriri when
attempting to follow it to the coast; it was set at intervals with
deadly spear pits, i.e. deep holes, the tops of which were masked
and the bottoms studded with firmly fixed, sharp-pointed spears—
pleasing contrivances arranged by the Maisina for the benefit of their
Doriri visitors. At length we emerged into solid country of jungle and
forest, and camped upon the bank of a narrow, rapid, and clear
river. I regret to say that, in his official report, Captain Barton
subsequently referred to my carriers as “crude savages of the
wildest kind!” They certainly did yell and dance, and indulge in mimic
warfare, half the night, until at my request they were rudely
thumped by either their chiefs or village constables; but that was
merely light-heartedness! Upon the following morning we resumed
our march, the constabulary now cutting our own track on a
compass line through heavy jungle and forest, until we came to a
river bed of some two hundred yards in width, down the middle of
which a rapid torrent flowed. This we forded by extending a long
light cotton rope, and all hanging on to it together, until the
expedition resembled a straggling long-legged centipede. Upon the
other side, we found our track-cutting much obstructed by masses of
fallen trees, that had been blown down by a whirlwind. In the early
afternoon, we struggled out of the tangle of timber on to the banks
of a watercourse, that was much wider than the last, and were here
told by the Maisina that we could not reach any further water before
night; we accordingly camped, in order to have a clear day in which
to cross the supposed waterless track. This statement afterwards
proved to be a lie on the part of the Maisina, who were beginning
bitterly to repent having been fools enough to consent to venture
near the Doriri, and wanted to prevent us from going any further. I
think though, that we should have been forced to camp in any case,
as Barton had developed some colicky pains in his tum-tum, which
later turned into a mild attack of dysentery.
The river we were camped upon, the Wakioki, is a most
extraordinary stream: its waters are of a greyish milky colour, and
highly charged with some fine substance which does not precipitate
when the water is allowed to stand; the consistency of the water
was that of thin treacle, and not that of water in which a man could
swim. A private slipped in his leg and foot, withdrawing them
immediately, and the water dried upon his skin like a coating of
whitewash. This was the point at which Sir Francis Winter was
deserted by the Maisina, in his attempt to reach and deal with the
Doriri. The country here was full of wild pigs, cassowary, wallaby,
and the enormous Goura pigeon, a bird nearly as big as a turkey;
duck and pigeon of all sorts were plentiful, and the Kaili Kaili carriers
spent a happy afternoon hunting. Grubs, snakes, pigs, etc., all were
game to them, and vanished down their ever-hungry gullets. The
Maisina hung about the camp, listening with apprehensive ears to
every distant sound. Two of the constabulary, who had gone
scouting in advance, returned at night and reported having
discovered fresh human footprints; these, the Maisina said, certainly
belonged to the Doriri, as no Collingwood Bay native would venture
so far inland; and, from the nearness to the coast, they thought the
Doriri must be bent on mischief.
Here was a pretty pickle! What were we to do? If we went
straight on, and there was a Doriri war party in the neighbourhood,
they would probably fall upon the Collingwood Bay villages, from
which we had drawn the best of the fighting men, and generally play
the devil, while we were laboriously wending our way to their
country. At last we decided to follow the footprints found by the
police; and, in the event of their leading us to a Doriri war party, fall
upon and destroy that party, or at all events drive it from the vicinity
of Collingwood Bay, before proceeding on our journey. Much of the
country here showed signs of extensive periodic inundation. Next
day we struck camp at dawn, and marched for the point at which
the police had found the footprints, Barton’s tum-tum being better,
having been treated with brandy, and lead and opium pills. Late in
the afternoon, after marching over rough, well-watered country, we
came to a stream running into a much larger one, and upon the
banks of which we discovered a freshly erected lean-to bush shelter,
such as are used by travelling natives, and a large number of newly
cut green boughs of trees, which had been used for making crude
weirs for catching fish. From the bush shelter, there led away in a
westerly direction—the direction of the land of the Doriri—a plainly
defined hunting track; this track we followed, until it was time to
camp for the night, finding everywhere signs of the recent prolonged
occupation by natives of the country through which we were
passing. As we pitched camp, we sent out some constabulary
scouts, and they returned after dark bringing with them some
burning fire sticks, and reported that upon the bank of the Wakioki
they had discovered some large lean-to shelters, only just vacated,
and with the cooking fires still burning in them.

CAPTAIN F. R. BARTON, C.M.G.

 Upon the following day we marched for this spot, and found the
shelter, as described by the police, situated at the junction of the
Buna and Wakioki Rivers. Here, by the size of the shelter and the
number of footprints, we came to the conclusion that it had
contained about thirty Doriri, who were probably attached to a much
larger party. We discovered here a curious and most ingenious
contrivance, in the shape of a litter, for conveying a sick or wounded
man. It consisted of a pole about eight feet long, passed through
three hoops or circles of rattan about two feet apart, the hoops
being thus suspended from the pole when carried on men’s
shoulders; round the inside of the circumference of the lower semi-
diameter of the circles or hoops, longitudinal strips or battens of
finely split palm were lashed, forming a soft and springy litter, upon
which an injured man could suffer very little from jolting on the
roughest track, or from out of which it was impossible to fall, or, with
any precaution at all on the part of the bearers, sustain any injury;
the central hoop was made to unfasten at the top, plainly as a
means of placing a man inside with least effort to himself. I have
made a rough sketch of the contrivance, which is decidedly superior
to any form of hand ambulance I have ever read of.

AAA. Carrying pole.

BB. Lathes of split palm,
CCC. Coir rope interlaced through lathes made to untie at
pole.

The Maisina now said that the Doriri had undoubtedly gone
down to the extensive sago swamps surrounding the Collingwood
Bay villages; but careful scouting, and full examination of the
direction of the Doriri footprints, which we now found to be very
numerous, all showed that they led up the Wakioki towards their
own country. We were now of the opinion that possibly the Doriri
had discovered our presence, and were retreating upon their own
villages; in any case, they were moving in that direction. Pursuit, and
that by forced marches, was now the order of the day. With far-flung
scouts, endeavouring to locate the Doriri ahead, we began the
chase, straining the endurance of the carriers to the last ounce; the
rear-guard of six constabulary and four village constables mercilessly
drove on the skulking Maisina, or helped the truly failing Kaili Kaili
with his load.
The bed of the Wakioki, up which we were now proceeding, is of
a most remarkable nature. It varies in width from 300 to 600 yards,
the banks being difficult to define, owing to the dense overgrowth of
young casuarina trees, through which many channels flow. Gaunt,
dead and dying casuarinas of huge size reared their enormous bulk
from the torn, boulder-strewn bed of the river; huge tree trunks and
lumps of wood, the bark stripped from them, and polished by eternal
friction, lay everywhere. In one place, where Mount MacGregor
descends to the river, the foot of the mountain was cut sheer off, as
though cleanly severed by the axe of some superhuman giant. It
was evident that the floods, which overwhelmed the country, fell as
rapidly as they rose, for light and heavy tree trunks were deposited
at every point, from the highest to the lowest; the fall of the
watercourse, where we first met it, was about one foot in two
hundred, and it increased in steadily growing gradient as we
ascended. We came to the conclusion (the right one, as I afterwards
ascertained on the second Doriri Expedition) that the floods and
inundations were due to enormous land-slips or avalanches,
comprised of hundreds of thousands of tons of rock, earth, and
timber, suddenly descending from Mount MacGregor into the narrow
gorge of the Wakioki, which skirted its spurs, thus blocking and
damming the river, until its growing weight and strength burst the
barriers and swept in one devastating wave over the lower country.
The colour and consistency of the river were due, I found out later,
to a wide stream of clayey substance, flowing from Mount
MacGregor, between rocky walls, into the river.
Early in the afternoon, we reached a point near the gorge from
which the Wakioki emerged; and there the track scrambled up a
loose boulder-strewn bank about thirty feet high, up which we
likewise clawed. Here we found, that though young casuarinas were
growing there, it yet bore signs, in the shape of boulders, drift-wood
and tree trunks, of being the bed of the river. We found many Doriri
shelters, that had only just been vacated, and still had the fires
burning in them. Here we pitched camp, right under the magnificent
Mount MacGregor, and gazed at the mountain pines on its spurs,
towering high above the surrounding tall forest trees. Our day had
been an interesting one: sometimes we were marching over huge
loose boulders, sometimes wading through a wet cream-cheesy sort
of pipe-clay, sometimes making our way over a hard-baked cement
of the same stuff, full of cracks, and throwing off a dry and
penetrating dust under our feet, which clogged our sweating skins
and choked our panting lungs; over all of which came the distant
angry voices of the likewise sweating rear-guard, as they
“encouraged” the labouring carriers to keep up with the column.
Shortly after our pitching camp, a violent thunder-storm rolled
down upon us from the mountains; streaks of vivid fork lightning
being succeeded by instantaneous claps of thunder, the whole being
followed by a torrential burst of rain; the river rose rapidly, and the
grinding roar of the enormous rolling boulders, swept before its
flood, made a din indescribable. The carriers whimpered with funk,
and I called in the sentries, feeling that that awful storm and night
were more than mortal man, standing at a solitary post, could be
expected to endure. I was also firmly convinced that no human
being, Doriri or otherwise, would be fool enough to be abroad on
such a night. We struck camp very early the next morning, only too
glad to get away from such a storm-torn, uncanny spot. After
marching a few miles, we found a Doriri track leaving the Wakioki,
and leading across the Didina ranges towards the Doriri country at
the head of the Musa River. The Maisina were now blue with funk,
and we greatly feared that they would bolt; but curses from us,
threats from the constabulary, and jeers from the Kaili Kaili, who told
them that if they left us, they (the Kaili Kaili) would make them the
laughing-stock of the coast as a set of women and weaklings, made
them pluck up their courage enough still to follow us. We found
growing on this track an extraordinary tough climbing bamboo, of a
vine-like nature, which, when cut with a knife, oozed from each joint
about a wineglassful of clear sweet water.
A severe march went on all day. Barton, who had now added a
very bad toothache to dysentery, was in command of the advance,
and feeling hard with his scouts for touch with the Doriri party
ahead; I was in charge of the rear-guard, and was severely driving
the fearful Maisina carriers. Night was closing in, the head of the line
had halted to camp, when back to me came an orderly, with a
message from Barton. “Hurry up; we are within touch of the Doriri.”
The Maisina, on hearing the magic word Doriri, rushed like scared
rabbits for the camp. Upon the rear-guard coming up with me,
Barton told me that the scouts ahead had seen a man up a tree,
who was calling to a party of Doriri ahead of him. The Maisina now
fairly collapsed with fright, and begged us to go back, saying that we
should all be eaten if we stayed. Barton and I consulted as to what
was to be done with them: to send them back was our best course,
but then, if by any remote chance there happened to be any Doriri
left in the country we had traversed, they would stand a good
chance of being cut to pieces, as we could not weaken our force, on
the eve of a fight, by detaching constabulary to escort them. They,
however, settled the question for themselves. Fearful as they were of
going on with us into the land of the dreaded Doriri, they were still
more afraid of leaving us and having to follow a lonely road back;
finding that we were determined to go on, and that the constabulary
and Kaili Kaili apparently treated the Doriri with contempt, they
quaveringly said they would follow.
We felled trees, and made our camp as strongly defensive as
possible; needless to say, the Maisina required no pressing to do
their share of this work, but toiled like veritable demons, clearing
scrub and dragging trees into a stockade, long after the order had
been given, “That will do the camp; post the night guard.”
Everything now pointed to the one conclusion, and that was that if
the party, on whose heels we had followed all the way from
Collingwood Bay, did not include the actual murderers by whom the
murders of six weeks ago had been committed, it undoubtedly
consisted of the tribe by whom innumerable murders had been done
previously, and who had kept a whole district in a state of tension
and misery for years. We were now right on the borders of the Doriri
country, for during the day we had ascended the summit of the
Didina Range, which formed the watershed between the streams of
Collingwood Bay and the Musa River. We had then crossed a fine
plateau and descended a small stream flowing towards the Musa,
which suddenly fell, by a series of cascades, over a precipice into a
valley; the track made a difficult circuit round this cascade, and
when we had descended into the valley we found the bottom
covered with stagnant water, forming a veritable quagmire,
impassable to our heavily laden men, although the Doriri had
somehow or other gone through it. Round this, we found it
necessary to cut a siding, which led us to the banks of the Ibinamu,
the most eastern affluent of the Musa River, which rose in Mount
MacGregor and was now seen by Europeans for the first time. The
Maisina guides had long since left the country with which they were
acquainted, and in any case would have been quite useless from
fright.
While in camp that night, Barton and I consulted together. There
appeared to us to be very little doubt, that the party just ahead of us
must be now quite aware of our presence in their vicinity, and be
laying their plans accordingly; as a matter of fact, we found out
afterwards that they were in a state of blissful ignorance. It never
for one moment entered the heads of the Doriri that any possible
danger could come to them from the cowed people of Collingwood
Bay, and Government or police they had merely heard of as a sort of
vague fable; of the effect of rifle fire they knew nothing, and with
spears they had never as yet met their match. “What are we going
to do now?” said Barton. “Capture or entirely destroy the party
ahead,” I replied. “I hate scientifically slaughtering unfortunate
savages, who are quite ignorant of a sense of wrongdoing,” said
Barton. “By every code in the world,” I said, “civilized or savage, the
people who commit wanton and unprovoked murder can expect
nothing else than to be killed themselves. Besides, our instructions
are plain and our duty clear.” The Maisina spent the night in a
miserable state of apprehension and fear, having quite made up their
minds that the cooking pots of the Doriri would be the ultimate fate
of the whole lot of us; the constabulary and Kaili Kaili were in a great
state of joy at the prospect of a fight, and the scroop-scrape of
stones on the edges of the Kaili Kaili tomahawks, the nervous
chatter of the Maisina, and restless prowling of the constabulary
went on all night. Poor Barton was writhing in agony from
toothache, and begged me to keep my “infernal savages” quiet; but
it was a hopeless task.
 ARMED CONSTABULARY, CAPE NELSON DETACHMENT

Dawn broke, and no time was lost in striking camp, and

resuming our march down the river in the direction of the voices
heard by the scouts on the previous day, and towards the Doriri
villages. Barton and I had an arrangement by which we took
alternate days in advance or rear, as the rear-guard work was
fatiguing and disagreeable in the extreme; on this day it happened
to be my turn in front. I saw plainly that unless something was done
soon to give the Maisina confidence in us, and in the power of the
constabulary to protect them, they would all knock up; they were
sick already from funk and want of sleep. First went the four scouts,
comprising two constabulary recruited from the Binandere people
and two village constables of the Kaili Kaili, hawk-eyed men, oiling
their way silently in advance, feeling for an ambush or touch with
the Doriri, and marking the track to be followed. Then I came, with
the advance-guard composed of my own men; next the Kaili Kaili,
then the Maisina, with village constables and constabulary scattered
at intervals among them, in order to hearten them; and last, Barton
and his police. The carriers had strict orders, in the event of fighting
in front, to rally on the rear-guard.
While a difficult piece of walking was causing the carriers to
straggle rather more than usual, and thus delaying Barton and the
rear-guard, two of the scouts came back and reported that they had
discovered men, how many they could not ascertain, in the bush on
one side of the river. These men were, in my opinion, the party
whom we had been following all along, with possibly others; and
from their silence, I concluded that they had either laid an ambush,
or still more probably formed a portion of a body of men coming
round on to the flank of our extended line. I dared not risk sending
the scouts out again, with a probability of their falling into the hands
of a strong party of Doriri, and should I delay to communicate with
Barton, and lose time in waiting for the rest of the police and
carriers to come up, I might allow time for an attack to develop on
our dangerously straggling line, with an absolute certainty of a
stampede on the part of the Maisina on top of Barton and the rear-
guard, and a possible bad slaughter before Barton knew what was
occurring or could clear his police. I therefore hastily detached seven
police; and ordered the others, with the village constables and Kaili
Kaili carriers who were nearest to the front, to draw out into the
clear river bed and there wait for the Commandant, who I knew
would be steadily coming up. In the meanwhile I, and my seven
men, made a detour into the scrub on the exposed side of our line,
with the object of both intercepting any attack that might be
coming, so as to allow of a better fighting formation being adopted,
and to come out on the rear and flank of the men seen by our
scouts.
After we had crawled and forced our way for some distance
under a dense tangled undergrowth over marshy ground, we
suddenly emerged upon a couple of bush shelters, from one of
which a Doriri sprang up in front of us with a frightful howl of
surprise and alarm, and armed with spear and club. In response to a
hasty order from me, the man was shot dead and a rush made upon
the shelters, from which three more men leaped, all armed. Two of
these men were at once knocked over by the police, and secured
uninjured; a fourth, who fought most desperately, frantically dashing
about with a club, leaped into the river, and though evidently
wounded in half a dozen places, still stuck to his club and made his
way across to the scrub on the opposite side of the river, hotly
pursued by two police. Never have I known a man so tenacious of
life as that Doriri. I myself sent four ·303 solid bullets through him
as he bolted, and yet he ran on. We found him afterwards dead in
the scrub, quite half a mile away. On gaining time to look round, I
saw about a dozen Kaili Kaili, who, in defiance of my order that they
were to remain on the river bed and wait for Barton, had thrown
down their loads and were rushing to join the two police chasing the
man across the river; while tearing, like devils possessed, through
the tangled undergrowth towards me came the remainder of the
Kaili Kaili and Mokoru, under the leadership of old Giwi’s son,
Mukawa. They afterwards explained that they were coming to the
help of the police and me. Knowing the awful job Barton must be
having to keep the Maisina together when the firing broke out
suddenly in front, and still expecting at any moment to see a rush of
Doriri on our now demoralized line, I recalled the police and
proceeded to collect carriers in the bed of the river, while Barton,
with the remaining carriers, was getting up to us.
 KAILI KAILI CARRIERS WITH THE DORIRI EXPEDITION

When Barton finally arrived, I found the poor old chap had
undergone a dreadful time. Firstly, his toothache had prevented him
from eating any breakfast; then, as he had painfully struggled over
the rough track shepherding the terror-stricken Maisina, the
roughness of the track and his empty condition had brought on a
recurrence of his dysentery. Halting, he had removed his revolver
and belts, and was in a helpless state, when suddenly the crack of
rifles came from the front, and his personal servant rushed at him
and endeavoured to buckle on his discarded accoutrements; the
Maisina were howling with terror and crowding all round him; his
constabulary, fairly foaming with impatience to be in the fight, were
endeavouring to make a break for me and took him all his time to
hold; while the Kaili Kaili threw all restraint to the winds, as they cast
their loads on the ground, and, flourishing their tomahawks, flew to
the sound of the firing. “Their own white master and their own
police” were fighting, that was enough for the Kaili Kaili; they should
not lack the assistance of their own people, be hanged to the Port
Moresby police! Kaili Kaili into the fighting line!
Three Dove Baruga men had accompanied the expedition as
carriers; they had been staying with the Kaili Kaili just before we
started, and, as they came from a village situated on the lower
Musa, the Doriri prisoners could understand their language;
therefore I used them as interpreters. The prisoners, upon being
questioned, said that they had formed a portion of a large party
returning from Collingwood Bay; and in response to a possibly not
quite fair question as to who had killed the Collingwood Bay people a
few weeks ago, they proudly said that they had themselves, or
rather the party to which they belonged. The remainder of them had
gone down the river to their village early that morning, and were
quite in ignorance of our presence in the valley. So accordingly we
started in pursuit.
The river bed had now widened to a bare boulder-strewn
watercourse, along which we could march in a close column instead
of the long straggling line of men in single file. About four in the
afternoon, during a period of intense still muggy heat, a rolling
crashing thunder-storm descended upon us from Mount MacGregor,
worse even than the last we had experienced. Fork and chain
lightning struck the boulders of the river bed, while balls of blue fire
rolled among them. “Better extend the men,” said Barton; “a close
column of men on the march gives off an emanation that is said to
attract lightning; and one of those flashes among our packed lot
might play hell.” I watched the course of the storm for a moment,
and then pointed out to Barton how the lightning only seemed to
strike among the boulders of the river bed, and not among the
forest trees bordering it. “I am all for camping in the tall timber,” I
said; “when the dry electrical disturbance has passed, the skies will
probably open and let go a veritable lake on top of us.” “It is said,”
remarked Barton, “that the neighbourhood of tall trees should be
avoided in a thunder-storm; but I’m hanged if I don’t think they are
safer than this place.” The Doriri prisoners were the only natives with
us at all apprehensive of the lightning, they knew the peculiar
beauties of their own storms, and were greatly relieved when they
found us wending our way to the trees; the Dove Baruga men had
by this time told them that we were a peculiar people, who did not
kill prisoners nor eat the bodies of the slain.
Before we were safely in camp, and during the operation of
pitching the tents, down came a torrential downpour of rain, soaking
us all to the skin. No one, who has not undergone the experience,
can possibly realize what a tropical rainstorm can be like; the water
does not fall in drops, but appears to be in continuous streams, the
thickness of lead pencils; it fairly bends one under its weight, and
half chokes one with its density; and all this in a steaming
atmosphere of heat that reduces one to the limpness of a dead and
decaying worm. In Captain Barton’s case, his misery was increased
by the spiky pangs of toothache and the slow gnawing of dysentery.
Tents were pitched at last, rain and storm passed, leaving a cool
and pleasant evening, camp fires burnt cheerily and cooks were busy
preparing the evening meal. Barton had stopped his toothache by
dint of holding his mouth full of raw whisky, and eased his tum-tum
with a prodigious dose of chlorodyne; pyjamas had replaced our
sodden clothing, the Kaili Kaili were gaily chattering, and even the
Maisina were plucking up their spirits, safe as they all thought in a
ring of watching sentries, when bang went a rifle some distance
away. I ran down to where a couple of sentries had been posted, at
the mouth of a stream leading into the camp; they had vanished. I
whistled for them, thinking that they had merely moved a few yards
away, and were concealed in the scrub; Barton heard my whistle,
thought that I wanted assistance, and came to me with a number of
constabulary. We then hastily dispatched half a dozen police to find
out what had become of the sentries; they did not return until after
dark, and then appeared bringing the missing men and another
private of constabulary with them. The latter bright individual had
quitted the camp without leave, and run into half a dozen Doriri, at
whom he had promptly fired; the Doriri decamped, as the sentries
deserted their posts and rushed to his assistance. The sentries were
told in chosen language exactly what was thought of them, and
fearful threats made as to the fate of the next men who left their
posts without orders. The roaming private was “punished,” as the
Official Report put it; as a matter of fact, he was soundly walloped
on the bare stern by his sergeant with a belt, a highly illegal but
most efficacious means of inducing him to see the error of his ways.
That night we had a little conversation with the Doriri prisoners,
and learnt that their villages were small and widely scattered, and
that their food supplies were none too good. They really made their
expeditions to Collingwood Bay in order to hunt game and make
sago, and the killing of the people there was only a supplementary
diversion, though of course the bodies of the slain gave them an
agreeable change of diet. “Will your people fight?” I asked. “Yes,”
was the reply, “of course they will; but those fire spears of yours are
dreadful things to meet. If it was the Maisina, now——” Here they
stared contemptuously at those unhappy people, who wilted
accordingly. “Never mind the Maisina, they are my people now,” I cut
in; “will the Doriri fight us?” “Yes, once,” was the reply, “until they
have learnt all about those fire spears.” “Yes, what then?” I queried.
“They will bolt for the hills, where you can’t find them, and starve
there, for we have little food.” “Monckton,” said Barton, “you are not
going to be callous brute enough to starve those unfortunate devils
in the hills?” “No,” I answered, “but I am going to break their
fighting strength, and teach them the futility of resisting a
Government order before I leave.”
The carriers now put in a request to me that they might be
allowed to eat any future Doriri killed; urging that, if they did so, it
would not only be a great satisfaction to them but also a
considerable saving to the stores of the expedition. “Really,” they
urged, “there was no sense in wasting good meat on account of a
foolish prejudice.” “You saw what happened to the disobedient
private to-day?” I said to them. “Yes, he was most painfully beaten
on the stern by the sergeant,” they said. “Quite so,” I replied. “Well,
the carrier, be he Kaili Kaili or Maisina, who as much as looks with a
hungry eye upon the body of a dead Doriri, will first be beaten in the
same way by the sergeant, then by the corporals and lance-
corporals, and then by the privates, until his stern is like unto the
jelly of baked sago.” This fearsome threat curbed the man-meat
hunger of the anthropophagi. After this we put in a peaceful and
undisturbed night; even the Maisina sleeping soundly, happy at last
in the belief that the dreaded Doriri would meet their match in the
constabulary, and that the chances of their going down Doriri gullets
were quite remote.
 CHAPTER XX

W E struck camp at daylight and moved down the river, soon

coming upon a number of well-built native lean-to
shelters, showing signs of having been recently and
hastily vacated; many articles of value to natives had been
abandoned, including some cleverly split slabs of green jade from
the hills of Collingwood Bay, which they used for making stone
heads for disc clubs, tomahawks or adzes; also earthenware cooking
pots, which the Maisina identified by the pattern as of their
manufacture. A little later we espied a small village situated upon a
spur of the Didina Range; a patrol of police searched the village, but
the inhabitants had decamped; a number of spears, however, were
taken and destroyed. Next we discovered, situated upon a rise in the
river bed, a village of about eighteen houses; this village was also
deserted, so we took possession and occupied it. In this village we
found ample evidence, in the shape of articles manufactured by the
Maisina and identified by them, of the complicity of its inhabitants in
the raiding; a large store also of recently manufactured sago, clearly
proved that they had only just returned from the Collingwood Bay
District.

Here we camped, in order to dry our clothes and give our

carriers a well-earned and much-needed rest. The prisoners told us
that the village was named Boure, and they looked on dismally while
the police and carriers slaughtered all the village pigs, and ravished
and devastated the gardens, which were but of small extent. Barton,
as he thought of the grief of the evicted inhabitants, looked quite as
unhappy as the prisoners, while the work of destruction went on,
and many a crack from his stick a too exultant yelling Kaili Kaili
received, if he incautiously approached too near that humanitarian.
“You know now what it feels like to have your villages raided,” said
the Dove Baruga to the prisoners; “we and the Maisina have had
years of it at your hands.” Our now happy carriers spent a cheery
night, gorging and snoring alternately, and well housed from the
rain.
Upon leaving Boure next morning, the track led down the river
bank through thick clumps of pampas-like grass, twelve feet high;
beastly dangerous country to traverse amongst a hostile people. I
was with the advance, when suddenly we heard the loud blowing of
war horns and the defiant shouting of a large force of men moving
up the river on our left. I at once changed our line of march towards
the direction of the Doriri, but after going on a short distance, the
grass became so thick and the track so narrow, as to prevent any
safe fighting formation being retained. A halt, therefore, was made,
and the constabulary formed into two bodies, fronting two lanes in
the tall grass, from either of which the now expected attack might
develop, the carriers being packed between the two lines of police.
The voices of Doriri calling, and horns blowing, could now be heard
on our front, rear, and, alternately, on each side, which looked as if
we were to receive an attack simultaneously on front, rear and
flanks. A worse position to defend it was almost impossible to
conceive: spearmen could approach unperceived, and launch their
spears, from the cover of the grass, into our packed men; while club
men could get right on top of us, before we could see to shoot with
any degree of certainty of hitting what we were shooting at; and
once amongst us, shooting would be out of the question for fear of
killing our own carriers. In the event of our advancing towards a
better position, we should be forced to straggle in a long line of
single file, which would expose our carriers to flank attack; and in
the case of a Doriri rush we should be in imminent danger of our line
being cut in two. The prisoners told us that the Doriri were now
shouting challenges and explaining that they were about to make an
end of the whole lot of us. We waited some time: the Maisina
whining and collapsing from funk, and the constabulary strung up to
the last pitch of nervous tension, waiting with finger on trigger for
the expected attack; one private, in his excitement, accidentally
exploding his rifle. I fancy that the Doriri were not quite certain of
our exact position, as we kept very quiet and the report of a rifle is
difficult to locate in thick cover, also I think they were no more
anxious to engage us in that horrible spot than we were anxious to
receive them there.
Barton and I consulted, for something had to be done, as the
Maisina were getting into a state of hysteria; we decided to bring
matters to a head by sending ten of the constabulary to crawl
through the grass and locate the Doriri, with a view to advancing
then our whole force. The ten men left, and shortly after yelled to us
to come on. Advancing, we found that the police had emerged from
the grass upon a long open stretch of sandy river bed, down which a
large body of armed natives were dancing towards them, yelling
furiously and brandishing spears, clubs and shields. The police were
standing in line, holding their fire for orders; I ran up to them, with
some additional police, and ordered them to fire into the advancing
natives. Crash went a volley, two men fell, shot dead, while many
others staggered into the surrounding long grass, more or less badly
wounded. The Doriri, though apparently frightfully surprised at the
effect of the rifle fire, still held their ground; but, as the steadily
firing constabulary line moved rapidly towards them, they began an
orderly retreat. Barton then came up; but, with a long line of
straggling carriers in the rear open to attack, we did not consider it
expedient to permit a police pursuit, and they were accordingly
recalled. We followed the tracks of the retreating party down the
Ibinamu, till it junctioned with the Adaua; here we found that the
greater portion of the attacking force had crossed to the other side
of the river.
The Maisina, from a state of utter collapse, had now ascended to
the highest pinnacle of jubilation; loud were their crows and great
their boasting. “The hitherto undefeated Doriri had met a force
comprising Maisina, and had retired before it with loss, and were
now in full retreat!” They made no allowance in their savage brains
for the fact that the unfortunate Doriri had encountered, for the first
time, a strange, powerful, and terrifying weapon in the shape of our
rifles—things which flashed fire, accompanied by a terrible noise,
and dealt death by invisible means at great distances. “I have never
known such damnable rotters as the Maisina,” said Barton, “they are
howling and paralysed with funk one minute, and gloating over a
few dead Doriri the next. They are like a costermonger rejoicing at a
victory over his wife or mother, gained by dint of kicking her in the
ribs.”
We now prepared to cross the river in pursuit of the retreating
Doriri: rafting was out of the question, as the river was eighty yards
wide, ran shoulder high, and was as swift as a mill-race. The first
thing to do was to place a piquet on the opposite bank to cover our
crossing; accordingly, some of the strongest swimmers amongst the
constabulary waded and swam across, with their rifles strapped on
their shoulders and cartridges tied on the tops of their heads, while
they were covered by watching men on our bank. Having crossed,
they yelled that there was a shallow bank in the middle of the river,
affording secure foothold; this information was a great relief to us,
as our cotton rope was not long enough to stretch across the full
river, and our lighter men (including Barton and myself) were not
strong enough to wade without its assistance. On that shoal,
therefore, we stationed some strong men, who held the end of our
rope; then we all crossed safely on to it, and there clung together,
until the constabulary, after repeated attempts, succeeded in
carrying the rope over the remainder of the river, where they tied it
to a tree. We then left our strongest men to hold on to the mid-river
end, and struggled across, with the loss only of a few bags of rice;
after which we hauled the rest of the men across, they clinging to
the end of the rope. Thus our crossing was accomplished.
Following the track of the retreating natives, we came to the
Domara River, where the Doriri foot-tracks dispersed in various
directions. The Domara had a fine wide sandy beach, admirable
country to fight in from our point of view. The prisoners now told us
that Domara village was close at hand, and there accordingly we
went, only to find it freshly deserted. It was a village containing, I
should estimate, about 180 to 200 men; it was circular in shape, and
surrounded by a moat, partly natural and partly artificial, ranging
from fifteen to twenty feet in width, and about ten feet in depth, and
clean and well kept. The houses were elevated on poles of from
twenty to thirty feet high; the poles were merely props, as the main
weight of the house was sustained by stout tree trunks, forming a
central king post; sometimes additional support was given by pieces
of timber fastened to live areca-nut palms. The village was certainly
an example of high barbaric engineering skill; moated as it was, and
with its high and easily defended houses, a very few of its male
inhabitants would be necessary for its defence against any force
armed only with spear and club. Hence it was easily seen how the
Doriri were enabled to keep so many men absent in Collingwood Bay
for so long a period. Some small gardens near were remorselessly
stripped to furnish the carriers with their evening meal, and every
village pig and dog was slaughtered; many spears and arms were
also found and burnt, the Maisina taking keen delight in cooking
Doriri pig over a fire made of Doriri spears. We remained two days in
this village, while patrols of police went out and endeavoured again
to get in touch with, or capture, Doriri; and the carriers plundered
and destroyed gardens to their hearts’ content and Barton’s grief.
The Doriri, however, had apparently had a bellyful of the awesome,
magic fire-spear, and had departed from their villages for the hills.
We found in the village, of all extraordinary articles, the brass chain
plate of a small vessel, now ground into an axe head.
Now evidently had come the time for departure: the Doriri had
learnt that there was a power stronger than themselves, and a
power, too, that could make itself unpleasantly evident. The most
essential thing to do was to convey a message to them, telling them
to abstain from raiding Collingwood Bay in the future, if they did not
wish again to incur the anger of that power. This we were shortly
able to do. We then left on our return journey, though by a different
route.
 Leaving Domara village we marched, for about five miles,
through jungle interspersed at intervals with small, old, and new
gardens; but nowhere did our scouts get into touch with the natives,
until we came to the Adaua again, near its confluence with the
Domara. The river, at this point, was about one hundred yards wide
and in flood, quite unfordable, and far too dangerous for rafts, as
the cataracts and rapids of the Musa, passing through the Didina
Range, were but a short distance below. The Doriri use a small,
triangular raft made of bamboo, and are much skilled in its use; our
men, however, were quite unable to manage the contrivance, it
requiring as much knack as a coracle. Ilimo village, to which one of
our prisoners belonged, was situated on a spur on the opposite
bank; and from thence we could hear the voices of natives calling to
one another as they watched our party. The scouts reported a small
village lower down the river, and upon the same bank, which our
prisoners told us was called Bare Bare; so there we went for the
night, or until the river went down sufficiently to permit of our
passage across. Bare Bare village was deserted, and apparently had
been so for some weeks; it was approached by narrow winding
tracts leading through a dense tall jungle of wild sugar-canes, which
were well sprinkled with spear pits. We cut a wide straight lane
through the jungle to the river, in order that our people might go
and come with water in safety. The scouts found near here a new
and much better ford than the one we had seen in the morning, and
which our lying prisoners had said was the only one.
Doriri yelled, howled, and blew horns on the opposite bank most
of the night, but did not venture to cross or interfere with us. In the
morning the scouts reported that the passage of the river was
possible at the new ford, so there we went. As we prepared to cross,
eight Doriri appeared on the opposite bank, in full war array,
dancing, yelling, turning and smacking their sterns at us. An
ominous sound of opening breech blocks spoke plainly of the opinion
the constabulary had formed of what would occur before we passed
the ford. “We must clear that bank of natives and place a guard
there, before the carriers attempt the river,” I said to Barton; “there
are only eight men in sight, but the scrub may swarm with them,
and if a man were swept off his feet by the current and carried down
the river, he would most certainly be speared before help could
reach him.” Barton agreed, and I ordered the six strongest of the
constabulary and a corporal to cross the river and guard the landing
point. The men started across, and had got within about fifty yards
of the dancing, yelling natives, who still defiantly remained there,
when I yelled to them: “Corporal, shoot those men!” The corporal
halted his men, and, shoulder high as they were in the fast-flowing
water, fell them into line; then slowly and deliberately, as if parading
at the butts, he put them through the movements of firing exercise.
“At one hundred yards with ball cartridge, load!” came his voice;
“ready!” “My God!” said Barton, “it is like witnessing an execution!”
and covered his face with his hands. “Present!” came the corporal’s
voice again; “fire!” One man leapt into the air and rolled over, some
of the others jumped as though stung; then they picked up the
fallen man and bolted into the scrub, while the constabulary
occupied the spot just vacated by them. “It is early in the morning,”
said Barton, “but I am going to have a little whisky after that.”
All that day and the next we spent in crossing some very steep
country in the Didina Range, in pouring rain, having awful difficulty
in starting fires with which to cook our food, as all the dead wood
was sodden with water. My personal servant, Toku, son of Giwi, at
last, however, found a species of tree, of which he had heard from
his father, that burnt readily even in its green state; after this we
always carried a supply of this tree with us, with which to start the
other wood. Getting fires lighted in rain on the mountains is not the
least of the minor discomforts of inland work in New Guinea, and
without fires one’s carriers are foodless, cold, and miserable. On
future expeditions, from the experience I gained on this one, I
always made my carriers make their carrying poles of a light, dry,
highly inflammable wood, and when the worst came to the worst,
took their poles to start the fires with, and made them cut fresh
green ones for use until we could again get light dry poles.
 Scrub itch and leeches made things very interesting for us in the
Didina hills. The former is a tiny little insect, almost invisible to the
naked eye, that falls in myriads like a shower from certain shrubs,
and promptly burrows under one’s skin; it is not until one is warm
under the blankets at night, that it gets its fine work in and renders
sleep impossible, until one collapses from exhaustion. Stinging trees
are another joy; they are harmless-looking shrubs with a pretty
glossy leaf, that sting one more than the worst of nettles; one of my
carriers, on the second Doriri expedition, fell over a bank into a
clump of the infernal things, and was in such agony that I had to put
him in irons to prevent him from destroying himself, while we
greased him all over with warm rifle oil. Leeches don’t need any
describing, only cursing, which they got very freely indeed from our
bare-legged police and carriers, as they beguiled their leisure
moments scraping festoons of the brutes off their legs; they
wriggled through one’s putties and breeches in a marvellous manner,
and rare indeed was the night when we did not find half a dozen
gorged brutes somewhere in our clothing, and knew that one would
later develop a like number of nasty little ulcers.
After crossing the Didina Range, we dropped down to a clear
stream, the Dudura, upon which was situated a village of the same
name; the inhabitants fled, but the constabulary succeeded in
catching one man and his wife. The Collingwood Bay carriers knew
of the village, both by name and reputation, and swore it was one of
the worst offenders in raiding them. I put a very unfair direct
question to the man. “Do you go to Maisina to kill people?” “Yes,” he
naïvely answered, “of course I do,” as if it was the most natural
thing in the world. “I am very sorry,” I told him, “but Government
disapproves of the promiscuous killing of people, and you must
come with me until you have learnt better.” The man’s wife was then
told that we were taking him away in order to complete his
education, but that later he would be safely returned to her. “You are
a set of murdering thieves,” she said. (She was, I may remark, a
strong-minded woman!) “I have not killed the Maisina, but you have
looted my house.” “Point out any man of ours, by whom you have
been robbed,” was the reply, as we ordered the whole expedition to
fall into line. Unerringly she picked out several of the Kaili Kaili,
incorrigible looters, and abused them vehemently, the while they
reluctantly made restitution. Her confidence was then gained by a
present of trade goods, to maintain her during the enforced absence
of her husband, and as payment for conveying from us to the Doriri
a full explanation as to the reason of our visit and hostility to them:
she was a most talkative dame, and I doubt not held forth at length
to the Doriri. Her husband seemed to regard the prospect of a
sojourn in gaol as rather a relief from the company of his very
masterful wife.
When we were leaving Dudura, Barton put in a plea for the
natives. “Monckton,” he said, “let us now avoid any conflict with the
natives; the poor devils did not know what they were doing in the
past, they have now had their warning, and I can no longer stand
seeing you use your police against them, coldly and mechanically, as
if they were a guillotine.” “All right, Barton,” I replied, “the rôle of
executioner does not appeal to me any more than it does to you, but
it is sometimes a necessary one; still, I will defer to your views, and
spare the people if possible. I only trust that the lesson we have
already read them has been sufficiently severe.” Afterwards I had
cause to repent my moderation, as the Doriri mistook our clemency,
as savages invariably do, for a sign of weakness, and went on the
raid again.
Taking our Dudura man with us and walking down the Dudura
stream, we soon emerged upon the banks of the Musa, which at this
point was a headlong tearing torrent, quite uncrossable; gradually,
as we descended the banks of the river, the valley widened and the
beach became better. In the afternoon, sounds of chopping were
heard, and a native was discovered busily engaged in felling a tree.
“I want that man alive and uninjured,” I said to the police. “He has
got an axe and looks a sturdy fellow,” they replied; “it looks difficult.”
Still, the constabulary, when told to do a thing, generally managed
to accomplish it, difficult or not. Four of them noiselessly slipped
away into the scrub, crept upon four sides and within a few yards of
the working man, unperceived by him; a private then attracted his
attention by yelling suddenly at him from behind; he gave a howl of
surprise and alarm, and sprang round to defend himself, with his axe
raised ready to strike. Then silently and swiftly as a springing
greyhound, a Mambare private rushed in and leapt upon his back,
bearing man and axe to the ground with the impetus of his rush; the
others sprang and threw themselves upon the pair, and after a
minute of a yelling, tangled, scrambling worry, during which he used
his teeth with good effect, our quarry was disarmed and handcuffed.
He was a fine, powerful, intelligent man, and, after he had been
induced to stop yelling and made to understand that he was not
going to be killed, he answered questions readily. “Who are you?”
we asked. “Gabadi, of the village of Dugari, lower down the Musa,”
he replied. The Maisina here said that Dugari was a most iniquitous
village, and concerned in all the raiding. It was, however,
imperatively necessary that we should get into friendly
communication with some of the tribes of the Upper Musa, and if we
retained Gabadi as a prisoner, we could not attain that end; we now
wished to make the object of our expedition clear beyond any
possibility of misconception in the minds of the Doriri. Gabadi was
therefore released, returned his axe, and given some tobacco, to
ease his mind of any feeling of fright or annoyance at the sudden
manner in which we had effected our introduction to him.
We then asked him to go down the river to his village, and tell
the people where we were, and that we wished to be friendly; also,
that we would buy all the food they chose to bring us. Gabadi said
that his wives had been in a camp some little distance away from
the place where we had caught him, and that they had fled while we
were engaged in making his acquaintance; he would therefore like
first to find them, in order to leave them safe in our camp while he
went off to Dugari. During his absence we pitched camp. After
howling for some time in the forest for his wives, he returned to us
in disgust; and, after remarking that the silly women would probably
alarm half the river, proceeded to make himself comfortable for the
night among the carriers. The intrusion of Gabadi was regarded by
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