Professional Documents
Culture Documents
Textbook Mathematical Modeling 1St Edition Christof Eck Ebook All Chapter PDF
Textbook Mathematical Modeling 1St Edition Christof Eck Ebook All Chapter PDF
Christof Eck
Visit to download the full and correct content document:
https://textbookfull.com/product/mathematical-modeling-1st-edition-christof-eck/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
https://textbookfull.com/product/introduction-to-mathematical-
modeling-1st-edition-mayer-humi/
https://textbookfull.com/product/mathematical-modeling-branching-
beyond-calculus-1st-edition-arangala/
https://textbookfull.com/product/mathematical-modeling-
applications-with-geogebra-1st-edition-jonas-hall/
https://textbookfull.com/product/mathematical-modeling-for-
business-analytics-1st-edition-william-p-fox/
Mathematical Modeling with Excel 2nd Edition Brian
Albright
https://textbookfull.com/product/mathematical-modeling-with-
excel-2nd-edition-brian-albright/
https://textbookfull.com/product/mathematical-modeling-of-
biosensors-2nd-edition-romas-baronas/
https://textbookfull.com/product/mathematical-modeling-of-
mitochondrial-swelling-messoud-efendiev/
https://textbookfull.com/product/affect-in-mathematical-modeling-
scott-a-chamberlin/
https://textbookfull.com/product/mathematical-modeling-
applications-with-geogebra-1st-edition-jonas-hall-thomas-
lingefjard/
Springer Undergraduate Mathematics Series
Christof Eck
Harald Garcke
Peter Knabner
Mathematical
Modeling
Springer Undergraduate Mathematics Series
Advisory Board
M.A.J. Chaplain, University of St. Andrews, St. Andrews, Scotland, UK
A. MacIntyre, Queen Mary University of London, London, England, UK
S. Scott, King’s College London, London, England, UK
N. Snashall, University of Leicester, Leicester, England, UK
E. Süli, University of Oxford, Oxford, England, UK
M.R. Tehranchi, University of Cambridge, Cambridge, England, UK
J.F. Toland, University of Cambridge, Cambridge, England, UK
Peter Knabner
Mathematical Modeling
123
Christof Eck Peter Knabner
Universität Stuttgart Department Mathematik
Stuttgart Universität Erlangen-Nürnberg
Germany Erlangen
Germany
Harald Garcke
Fakultät für Mathematik
Universität Regensburg
Regensburg
Germany
Mathematics Subject Classification (2010): 00A71, 34-01, 35-01, 49-01, 74-01, 76-01, 80-01
1
See the preface to the German edition for more detailed suggestions for the use of this textbook.
vii
viii Preface
It fills us with deep sadness that our scholar, colleague, and friend Christof Eck
could not participate in this enterprise. Extremely untimely, just being 43 years of
age, he passed away after a long illness. This volume is dedicated to his memory.
which students and lecturers of mathematics need in order to understand models for
problems in the sciences and engineering and also to derive them. On the other hand,
this book contains a variety of interesting, practically relevant examples for the
mathematical theories often only experienced at an abstract level during the study of
mathematics and thus answers the question often posed “what do I need this for?”.
While it cannot substitute any of the textbooks dealing with the underlying mathe-
matical structures such as linear systems of equations/linear algebra or ordinary or
partial differential equations, it nevertheless contains essential aspects of the analysis
of the models. One aim in particular is to illustrate the interactions between mathematics
and applications, which unfortunately are often neglected in mathematics courses.
Furthermore, this textbook also addresses students from the sciences and from
engineering and offers them an introduction into the methods of applied mathe-
matics and mechanics.
The content of this book is restricted to deterministic models with continuous
scales, as they are in the center of classical natural sciences and engineering. In
particular, stochastic models are beyond the scope of this book, and the same
applies to processes at very small scales for which particular models or models from
quantum mechanics and its approximations are feasible. Also, models from eco-
nomics are not in the focus of this book, as stochastic approaches play an important
role there.
An essential concept of this book consists in using the mathematical structures
(and the knowledge about them) as an ordering principle and not the different fields
of applications. This reflects the strength of mathematics, lying in the fact that one
concept can be used for totally different problem classes and fields of applications.
It allows dealing with examples from different fields of applications efficiently
without being forced to always repeat the same mathematical basic structures: This
line will be followed in Chapter 1 and in Chapters 2, 4, 6, and 7. In this order, one
finds embedded Chapters 3 (thermodynamics) and 5 (continuum mechanics). They
provide the necessary links to the natural sciences and engineering. Of course, these
chapters are also shaped by the application of mathematical tools.
The restriction of the subjects at the level of application corresponds to a
restriction at a mathematical level: Throughout this book, we use knowledge from
linear algebra and analysis intensively. Chapter 4 relies on the knowledge provided
by courses in analysis or in ordinary differential equations, and Chapter 5 makes use
in an essential way of the methods of the multidimensional differentiation and
integration (integral theorems) and in this way from the more advanced aspects of
analysis. In Chapter 7, the foundations of the geometry of curves and surfaces play
an important role. It is impossible to define a clear delimitation to the analysis of
partial differential equations. Knowledge from this field and also from linear
functional analysis certainly is useful for Chapter 6, but not necessary.
A discussion of mathematical results about partial differential equations takes
place in this chapter only insofar as there is a tight linkage to the model interpre-
tation. Therefore, the presentation cannot be completely rigorous, but possible gaps
and necessary consolidations are pointed out. In this way, this chapter does not
necessarily require an intense study of the analysis of partial differential equations,
x Preface
author at the University of Regensburg and by the first and third author at the
University of Erlangen several times, and therefore, it is the result of a complex
developing process. During this process, the authors received important support.
The authors express their thanks to Bernd Ammann, Luise Blank, Wolfgang
Dreyer, Michael Hinze, and Willi Merz for valuable suggestions. Sincere thanks are
given to Barbara Niethammer, who together with the second author has lectured a
course on mathematical modeling at the University of Bonn from which much
material entered the present book. For careful proofreading, we thank Martin Butz,
Daniel Depner, Günther Grün, Robert Haas, Simon Jörres, Fabian Klingbeil, David
Kwak, Boris Nowak, Andre Oppitz, Alexander Prechtel, and Björn Stinner. In TEX
writing, we were supported by Mrs. Silke Berghof and in particular by Mrs. Eva
Rütz who typed a large part of the manuscript and worked on the numerous figures
with strong dedication—to both our cordial thanks. We thank Serge Kräutle who
provided the figure at the cover—a numerical simulation of the Kármán vortex
street. We would also like to thank cordially Ulrich Weikard for providing
Figure 6.14 and James D. Murray for providing Figure 6.10.
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 1
1.1 What Do We Mean by (Mathematical) Modeling? . . . . . . . . .... 1
1.2 Aspects of Mathematical Modeling: Example of Population
Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Population Models with Restricted Resources. . . . . . . . . . . . . . . . 8
1.4 Dimensional Analysis and Scaling . . . . . . . . . . . . . . . . . . . . . . . . 12
1.5 Asymptotic Expansions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.6 Applications from Fluid Mechanics . . . . . . . . . . . . . . . . . . . . . . . 27
1.7 Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.8 Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2 Systems of Linear Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.1 Electrical Networks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.2 Space Frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.3 Constrained Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
2.4 Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.5 Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3 Basic Principles of Thermodynamics . . . . . . . . . . . . . . . . . . . . . . .... 75
3.1 The Model of an Ideal Gas and the Maxwell–Boltzmann
Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 76
3.2 Thermodynamic Systems and the Thermodynamic
Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 80
3.3 The First Law of Thermodynamics . . . . . . . . . . . . . . . . . . . . .... 81
3.4 The Second Law of Thermodynamics and the Notion
of Entropy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.5 Thermodynamic Potentials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
3.6 The Legendre Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
3.7 The Calculus of Differential Forms . . . . . . . . . . . . . . . . . . . . . . . . 99
3.8 Thermodynamics of Mixtures and the Chemical Potential . . . . . . 102
xiii
xiv Contents
are measured and therefore afflicted with measurement errors. Furthermore in this
example certainly the gravitational force of the Earth has to be taken into account, but
its dependence on the flight altitude can be neglected. In a similar way the influence
of the rotation of the Earth can be neglected. On the other hand the influence of air
resistance cannot be neglected. The negligible effects are exactly those which make
the model equations more complex and require additional data, but do not improve
the accuracy of the results significantly.
In deriving a model one should make oneself clear what is the question to be
answered and which effects are of importance and have to be taken into account
in any case and which effects are possibly negligible. The aim of the modeling
therefore plays a decisive role. For example the model assumptions mentioned above
are sensible for the flight trajectory of a soccer ball, but certainly not for the flight
trajectory of a rocket in an orbit around the Earth. Another aspect shows the following
example from weather forecasting: An exact model to compute the future weather for
the next seven days from the data of today cannot serve for the purposes of weather
forecast if the numerical solution of this model would need nine days of computing
time of the strongest available supercomputer. Therefore often a balance between the
accuracy required for the predictions of a model and the costs to achieve a solution
is necessary. The costs can be measured for example by the time which is necessary
to achieve a solution of the model and for numerical solutions also by the necessary
computer capacities. Thus at least in industrial applications costs often mean financial
costs. Because of these reasons there can be no clear separation between correct or
false models, a given model can be sensible for certain applications and aims but not
sensible for others.
An important question in the construction of models is: Does the mathematical
structure of a model change by neglecting certain terms? For example in the initial
value problem
ε y (x) + y(x) = 0 , y(0) = 1
with the small parameter ε one could think about omitting the term εy . However,
this would lead to an obviously unsolvable algebraic system of equations
y(x) = 0 , y(0) = 1 .
The term neglected is decisive for the mathematical structure of the problem indepen-
dent of the smallness of parameter ε. Therefore sometimes terms which are identified
as small, cannot be neglected. Hence, constructing a good mathematical model also
means to take aspects of analysis (well-posedness) and numerics (costs) of the model
into account.
The essential ingredients of a mathematical model are
x(tn ) = r n x(t0 ) .
with a factor p still unknown. From r = 1.15 for Δt = 1 year we conclude that p =
0.15/year. Taking this for granted then for Δt = 2 years one has r = 1.3. Therefore
after 6 years, (6 = 3 times 2) the farmer has
or more precisely
x(t + Δt) − x(t)
lim = p x(t) ,
Δt→0 Δt
i.e.,
x (t) = p x(t) . (1.2)
If the data are as above, i.e., a time increment of Δt = 1 year and a growth rate
r = 1.15, this means
e p·1 year = 1.15
and therefore
p = ln(1.15)/year ≈ 0.1398/year .
this is (1.1) with r = 1 + Δt p. In the case p < 0 a time increment of the size
Δt < (− p)−1 has to be chosen to achieve a sensible sequence of numbers. On the
other hand using the implicit Euler method one gets
i.e., (1.1) with r = (1 − Δt p)−1 . Here for p > 0 the time step has to be chosen such
that Δt < p −1 . By Taylor series expansion one can see that the different growth
factors coincide for small Δt “up to an error of the order O (Δt)2 ”:
(1 − Δt p)−1 = 1 + Δt p + O (Δt p)2 .
The connection between the continuous and the discrete model therefore can be
established by an analysis of the convergence properties of the numerical method.
For the (explicit or implicit) Euler method one gets for example
|x(ti ) − xi | ≤ C(te ) Δt ,
6 1 Introduction
where x(ti ) is the exact solution of (1.2) at time ti and xi is the approximate solution
of the numerical method, assuming that ti ≤ te , where te is the given final time for the
model. For details about the analysis of numerical methods for ordinary differential
equations we refer to the textbooks of Stoer and Bulirsch [123] and Deuflhard and
Bornemann [28].
Both models, the discrete and the continuous, have the seeming disadvantage
that they also allow non-integer solutions, which obviously are not realistic for the
considered example. The model describes — as it is true for every other model —
not the total reality but only leads to an idealized picture. For small populations
the model is not very precise, as in general population growth also depends heavily
on stochastic effects and therefore cannot be computed precisely in a deterministic
way. In addition for small populations the model assumptions are questionable, in
particular one neglects the age and the sex of the animals. In the extreme case of a
herd of two animals obviously the growth will depend heavily on the fact whether
there is a male and a female animal, or not. For large populations on the other hand
one can assume with a certain qualification that it possesses a characteristic uniform
distribution in age and in sex, such that the assumption of a growth proportional to
population size make sense.
The substitution of integer values by real numbers reflects the inaccuracy of the
model. Therefore it is not sensible to change the model such that integer values in
the solutions are enforced. This would only lead to an unrealistic perception of high
accuracy of the model. For a small population a stochastic model, which then “only”
provides statements about the probability distribution of the population size, makes
sense instead of deterministic models.
Nondimensionalization
The quantities in a mathematical model generally have a physical dimension. In the
population model (1.2) we have the units number and time. We denote the physical
dimension of a quantity f with [ f ] and abbreviate the units number of entities by A
and time by T . Therefore we have
[t] = T ,
[x(t)] = A ,
A
[x (t)] = ,
T
1
[ p] = .
T
The specification of a physical dimension is not yet a decision about the physical unit
of measurements. As a unit of measurement for time one can use seconds, minutes,
hours, days, weeks, or years, for example. If we measure time in years, then t is
indicated in years, x(t) by a number, x (t) in number/years and p in number/years.
To get models as simple as possible and furthermore in order to determine char-
acteristic quantities in a model, one can nondimensionalize the model equations.
1.2 Aspects of Mathematical Modeling: Example of Population Dynamics 7
For this aim one defines a characteristic value for every appearing dimension and
correspondingly a unit of measurement. Here it is not necessary to choose one of the
common units as for example seconds or hours but it is more appropriate to choose
a unit adapted to the problem. For the population model there are two dimensions,
therefore two characteristic values are needed, the characteristic number x and the
characteristic time t. These are chosen in such a way that the initial data t0 and
x0 = x(t0 ) are as simple as possible. Therefore a convenient unit of measurement
for time is given by
t − t0
τ= ,
t
where t denotes a unit of time which still has to be specified, and as a unit for number
we choose
x = x0 .
Setting
x
y=
x
and expressing y as a function of τ ,
x(tτ + t0 )
y(τ ) = ,
x
one obtains
t
y (τ ) = x (t)
x
and therefore the model becomes
x
y (τ ) = p x y(τ ) .
t
This model gets its most simple form for the choice
1
t= . (1.3)
p
y (τ ) = y(τ ) ,
y(0) = 1 . (1.4)
From this solution all solutions of the original model (1.2) can be achieved by using
a transformation:
t = p n x0m with n, m ∈ Z .
The only possible solution of this equation is given by n = −1, m = 0, if the number
of animals is interpreted as a dimension of its own. Thus we get exactly (1.3).
In more complex models typically the model cannot be reduced to a single problem
by nondimensionalization but the number of relevant parameters can be strongly
reduced and the characteristic parameters can be identified. This also relates to the
corresponding experiments: For instance, from the nondimensionalization of the
equations for airflows one can conclude how the circulation around an airplane can be
experimentally measured by using a (physical) model for the airplane much smaller
in scale. We will explain dimensional analysis in one of the following sections using
a more meaningful example.
For large populations in nature a constant growth rate is not realistic anymore. A
restriction of the habitat, or the available nutrients, or other mechanisms impose
limitations on the growth. To construct a model it is feasible for such situations to
assume that there is a certain capacity x M > 0 for which the resources of the habitat
are still sufficient. For population quantities x smaller than x M the population still
can grow, but for values larger than x M the population decreases. This means that
the growth rate p now depends on the population x, p = p(x), and that
1.3 Population Models with Restricted Resources 9
have to hold true. The most simple functional form satisfying these conditions is
given by a linear ansatz for p, i.e.,
with a parameter q > 0. With this ansatz we obtain the differential equation
as a model. The additional term −q x(t)2 is proportional to the probability for the
number of encounters of two specimens of the population per unit of time. The
term represents the more competitive situation if the population size increases, the
so-called “social friction”. The Eq. (1.5) has been proposed by the Dutch biomath-
ematician Verhulst and is called logistic differential equation or equation of limited
growth.
Equation (1.5) also can be solved in closed form (compare Exercise 1.3). From
x
=q
x(x M − x)
and by integration
ln(x(t)) − ln |x M − x(t)| = x M qt + c1 , c1 ∈ R .
x(t)
= c2 e x M qt ,
x M − x(t)
and
c2 x M e x M qt xM
x(t) = = .
1 + c2 e x M qt 1 + c3 e−x M qt
From this exact solution the following properties can be easily derived:
• If x0 is positive, the solution always stays positive.
• If x0 is positive, then for t → +∞ the solution converges to the equilibrium point
x∞ = x M .
The graph of x can be sketched also without knowing the exact solution. From (1.5)
first we conclude
x > 0 , if x < x M ,
x < 0 , if x > x M .
Furthermore we have
x = (x ) = (q (x M − x) x) = q(x M − x) x − q x x
= q(x M − 2x)x = q 2 (x M − 2x)(x M − x) x .
Thus the solution curves have an inflection point at x M /2 and the curves are concave
in the interval between x M /2 and x M , and convex otherwise. Solutions of the logistic
differential equation are depicted in Fig. 1.1.
Stationary Solutions
For more complex time-depending models a closed form solution often cannot be
found. Then it is useful to identify time independent solutions. Such solutions can
be computed using the time dependent model by just setting all time derivatives to
zero. For our model with restricted growth one gets
0 = q xM x − q x2 .
xM
xM
2
x0 = 0 and x1 = x M .
These are the solutions of the original model for specific initial data. Often time
independent solutions appear as so-called stationary limits of arbitrary solutions
for large times, meaning that they are solutions constant in time towards which
time dependent solutions converge for large times. Typically this only appears if the
stationary solution is stable in the following sense: If the initial data is only changed
slightly then also the solution changes only slightly. Using the exact solution (1.6)
the question of stability can be easily answered for the logistic differential equation:
The solution for the initial value
x(t0 ) = ε
x M (x M + ε)
xε (t) = ,
(x M + ε) − ε e−x M q(t−t0 )
one can conclude also without knowing the exact solution that the distance to x M
can only decrease for increasing time as from xε (t) > x M it follows xε (t) < 0 and
from xε (t) < x M it follows xε (t) > 0. Therefore the stationary solution x M is stable.
Stability is of importance, as in nature in general no instable stationary solution can
be observed, therefore they are irrelevant for most practical applications. For more
complex models sometimes no closed form solution for the time dependent equation
can be derived. However, there are techniques of stability analysis, with which often
the stability properties of stationary solutions can be deduced. Often this is done
by means of a linearization of the problem at the stationary solution followed by a
computation of the eigenvalues of the linearized problem. This will be explained in
more detail in Chap. 4.
12 1 Introduction
F
a= ,
m
where a denotes the acceleration of the body and F the force acting on the body.
This force is described by Newton’s law of gravitation
mE m
F = −G ,
(x + R)2
where G ≈ 6.674 · 10−11 N · m2 /kg2 denotes the gravitational constant, m E the mass
of the planet, R the radius of the planet and x the height of the body, measured from
the surface of the planet. We neglect the air resistance in the atmosphere and consider
the planet to be a sphere. If one defines the constant g by
Gm E
g= ,
R2
one gets
g R2m
F =− .
(x + R)2
For the Earth we have g = 9.80665 m/s2 , the gravitational acceleration. The motion
of the body then is described by the differential equation
g R2
x (t) = − . (1.7)
(x(t) + R)2
x(0) = 0 , x (0) = v 0 ,
Π = v 0a g b R c ,
which are either dimensionless (case (i)), or have the dimension of a length (case
(ii)), or have the dimension of a time (case (iii)). From
a b
L L
[Π ] = L c = L a+b+c T −a−2b
T T2
it follows:
v 02
ε= (1.8)
gR
R −b
τ = v 0−1−2b g b R b+1 = ε .
v0
We will now try to nondimensionalize Eq. (1.7). To this purpose we consider a unit
for length x and a unit for time t and represent x(t) in the form
14 1 Introduction
x(t) = x y(t/t) .
i.e.,
x 1
y (τ ) = − . (1.9)
2
t g ((x/R) y(τ ) + 1)2
t
y(0) = 0 and y (0) = v0 .
x
Now we want to choose x and t in such a way that as many of the appearing parameters
as possible equal 1. However, here we have more parameters than scaling units,
namely the three parameters
x x t
2
, , and v0 .
t g R x
Hence, only two of these parameters can be transformed to one and therefore there
are three different possibilities:
x x R
(a) 2 = 1 and = 1 are a consequence of x = R, t = , then the parameter
t g R g
t v0 √
is given by v 0 = √ = ε using ε from (1.8). Therefore the model reduces
x Rg
to
1 √
y (τ ) = − , y(0) = 0 , y (0) = ε . (1.10)
(y(τ ) + 1)2
x t R x
(b) = 1 and v 0 = 1 can be deduced from x = R and t = , leading to 2 =
R x v0 t g
v 02
= ε for the third parameter. Then the dimensionless model is given by
Rg
1
ε y (τ ) = − , y(0) = 0 , y (0) = 1 .
(y(τ ) + 1)2
x t v0 v2
(c) = 1 and v 0 = 1 are a consequence of t = and x = 0 . Then the third
2
t g x g g
x v 02
parameter is given by = = ε. Thus the dimensionless model reads
R gR
1.4 Dimensional Analysis and Scaling 15
1
y (τ ) = − , y(0) = 0 , y (0) = 1 . (1.11)
(ε y(τ ) + 1)2
Let us mention that there is a fourth possibility. Besides (1.9) we may also use the
equivalent formulation
x3 1
y (τ ) = − .
t
2
g R2 (y(τ ) + R/x)2
1
y (τ ) = − , y(0) = 0 , y (0) = 1
(y(τ ) + ε)2
v 02
ε= = 10−6 .
Rg
1
y (τ ) = − , y(0) = 0 , y (0) = 0 .
(y(τ ) + 1)2
Because of y (0) < 0 and y (0) = 0 this model leads to negative solutions and there-
fore it is extremely inexact and of no use. The reason lies in the scaling within the
nondimensionalization: The parameters t and x here are given by
R
t= = 103 s and x = 107 m ,
g
both scales are much too large for the problem under investigation. The maximal
height to be reached and the instance of time for which it is reached are much
smaller than the scales x for length and t for time and therefore are “hardly visible”
in the nondimensionalized model.
The model (b) reduces to
1
0=− , y(0) = 0, y (0) = 1 .
(y(τ ) + 1)2
Another random document with
no related content on Scribd:
{27}
In his opinion the Ashantis had been preparing ever since the
British occupation in 1896 to reassert their independence."
The Governor was, himself, in Kumassi when the Ashantis first
attacked it, on the 25th of March, and he has given an account
of the desperate position in which the few British officials,
with their small native garrison and the refugees whom they
tried to protect, were placed. "Our force," said Sir Frederic
Hodgson, "consisted of only some 200 Hausas, while there is
reason to believe that we had not less than 15,000 Ashantis
surrounding us. In addition to our own force we had to protect
some 3,500 refugees, chiefly Mahomedan traders, Fantis, and loyal
Kumassis, none of whom we were able to take into the fort,
where every available bit of space was required for military
purposes. It was heartrending to see the efforts of these poor
people to scale the walls or break through the gate of the
fort, and we had to withdraw the Hausas from the cantonments
and draw a cordon round the refugees. It is impossible to
describe the horror of the situation with these 3,500 wretched
people huddled together without shelter under the walls of the
fort. That same night a tornado broke over Kumassi, and the scene
next morning with over 200 children was too terrible for
words. Afterwards they were able to arrange shelters for
themselves." Near the end of April, two small reinforcements
from other posts reached Kumassi; but while this strengthened
the numbers for defence, it weakened the food supply. Taking
stock of their food, the besieged decided that they could hold
out until June 23, and that if the main body then marched out,
to cut, if possible, their way through the enemy, leaving a
hundred men behind, the latter might keep the fort until July
15. This, accordingly, was done. On the 23d of June Governor
Hodgson, with all but 100 men, stole away from Kumassi, by a
road which the Ashantis had not guarded, and succeeded in
reaching the coast, undergoing great hardships and dangers in
the march. Meantime, an expedition from Cape Coast Castle was
being energetically prepared by Colonel Sir J. Willcocks, who
overcame immense difficulties and fought his way into Kumassi
on Ju]y 15, the very day on which the food-supply of the
little garrison was expected to give out. The following
account of his entry into Kumassi is from Colonel Willcocks'
official report: "Forming up in the main road, we marched
towards Kumassi, a mile distant, the troops cheering wildly
for the Queen and then followed silence. No sound came from
the direction of the fort, which you cannot see till quite
close. For a moment the hideous desolation and silence, the
headless bodies lying everywhere, the sickening smell, &c.,
almost made one shudder to think what no one dared to
utter—'Has Kumassi fallen? Are we too late?' Then a bugle
sound caught the ear—'the general salute'—the tops of the
towers appeared, and again every man in the column, white and
black, broke into cheers long sustained. The brave defenders
had at last seen us; they knew for hours' past from the firing
growing ever nearer that we were coming, yet they dared not
open their only gate; they perforce must wait, for even as we
appeared the enemy were making their last efforts to destroy
the outlying buildings, and were actually setting them on fire
until after dark, when a party of 100 men went out and treated
them to volleys and cleared them out. If I have gone too fully
into details of the final scene, the occasion was one that
every white man felt for him comes perhaps but once, and no
one would have missed it for a kingdom."
ASSASSINATIONS:
Of President Barrios.
Of President Borda.
Of Governor Goebel.
Of President Heureaux.
Of King Humbert.
Of Professor Mihaileano.
Of M. Stambouloff.
ATHENS: A. D. 1896.
The revival of Olympic games.
{28}
ATLANTA: A. D. 1895.
The Cotton States and International Exposition.
AUSGLEICH, The.
----------AUSTRALIA. Start--------
AUSTRALIA:
Recent extensions of Democracy in the Australian Colonies
and New Zealand.
Social experiments.
"One of the most hopeful signs of the day is that, with the
help of the representatives of labor in Parliament, Australian
governments have done much within recent years to mitigate the
excess of population in the large towns, and to replace the
unemployed upon the land. Of course mistakes have been made.
In some cases settlers have failed through lack of
agricultural knowledge; in others, on account of the
barrenness of the soil. In South Australia, the village
settlements, which were avowedly started as an alternative to
relief works, have been only a modified success. In New
Zealand, village settlements have produced very satisfactory
results. … In Victoria, a labor colony has been established,
with the entire support of the trades-unionists, to which the
unemployed may be sent, and at which they receive, at a very
low rate of wages, a course of instruction in agricultural
pursuits which enables them subsequently to obtain private
employment with farmers or others. In New Zealand, I found a
very strong feeling among trades-unionists that it would be to
the interest of the workingmen themselves if a penal colony
were established, on the lines of those which exist in
Germany, to which loafers might be sent, and at which they
would be compelled to work, with the alternative of
starvation."
H. De R. Walker,
Australasian Extensions of Democracy
(Atlantic Monthly, May, 1899).
"Nor does the ominous kind of hint that preceded the Jameson
Raid fail to be uttered. Only three terminations, according to
Dr. Thurston, are possible if Sir John Forrest does not modify
his Krugerite policy towards the mining settlers:
'(1) Separation of the goldfields.—This would be only fair to
the goldfields; but thousands of Outlanders have settled in
the other parts of the colony, and this step would not redress
their wrongs. The practical result of this step would be
prosperity for the goldfields, but almost ruin for the rest of
the colony.
(2) Revolution.—I fear this is much more probable than is
generally thought. Unless a material change takes place
quickly there will be bloodshed in this colony.
(3) General depression, practically equivalent to bankruptcy.'
Separation, however, and the creation of a new colony, which
would include the Goldfields district and come down to the
sea, and would immediately join the Australian Federation, is
the remedy officially proposed by the representatives of the
Outlanders. …
{30}
S. Low,
Enigmas of Empire
(Nineteenth Century, June, 1900).
"It may be well here to remark that there are a few articles,
notably raw sugar, glucose, molasses, and treacle, upon which
the duty will be removed gradually, so as not to wantonly
disturb vested interests, but, with these exceptions, the
change is a very sweeping one."
AUSTRALIA: A. D. 1897.
Conference of colonial premiers with
the British Colonial Secretary.
AUSTRALIA: A. D. 1900.
Federation of the Australian Colonies.
The steps by which the Union was accomplished.
Passage of the "Commonwealth of Australia Constitution Act" by
the Imperial Parliament.
W. G. Beach,
The Australian Federal Constitution
(Political Science Quarterly, December, 1899).
AUSTRALIA: A. D. 1900.
The question of the Federal Capital.
Majority. 261,623