Full download ebook at ebookname.

com

Solutions Manual for Introduction to Mathematical

Modeling and Chaotic Dynamics 1st Edition Kumar
Upadhyay

https://ebookname.com/product/solutions-manual-for-
introduction-to-mathematical-modeling-and-chaotic-
dynamics-1st-edition-kumar-upadhyay/

OR CLICK BUTTON

DOWLOAD NOW

Download more ebook from https://ebookname.com

More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Introduction to Algebra Solutions Manual 2nd Edition

Richard Rusczyk

https://ebookname.com/product/introduction-to-algebra-solutions-
manual-2nd-edition-richard-rusczyk/

Chaotic Dynamics Tama´S Te´L

https://ebookname.com/product/chaotic-dynamics-tamas-tel/

Instructor Solutions Manual for Analysis With an

Introduction to Proof 4th Edition Steven R. Lay

https://ebookname.com/product/instructor-solutions-manual-for-
analysis-with-an-introduction-to-proof-4th-edition-steven-r-lay/

Chaotic Dynamics Fractals Tilings and Substitutions 1st

Edition Geoffrey R. Goodson

https://ebookname.com/product/chaotic-dynamics-fractals-tilings-
and-substitutions-1st-edition-geoffrey-r-goodson/
Chaotic dynamics of nonlinear systems S. Neil Rasband

https://ebookname.com/product/chaotic-dynamics-of-nonlinear-
systems-s-neil-rasband/

Mathematical Modeling of Physical Systems An

Introduction Engineering Technology 1st Edition
Basmadjian

https://ebookname.com/product/mathematical-modeling-of-physical-
systems-an-introduction-engineering-technology-1st-edition-
basmadjian/

Logic and Discrete Mathematics A Concise Introduction

Solutions Manual 1st Edition Willem Conradie

https://ebookname.com/product/logic-and-discrete-mathematics-a-
concise-introduction-solutions-manual-1st-edition-willem-
conradie/

Modeling Carbon and Nitrogen Dynamics for Soil

Management 1st Edition M.J. Shaffer

https://ebookname.com/product/modeling-carbon-and-nitrogen-
dynamics-for-soil-management-1st-edition-m-j-shaffer/

Introduction To Mathematical Logic 1st Edition Michal

Walicki

https://ebookname.com/product/introduction-to-mathematical-
logic-1st-edition-michal-walicki/
SOLUTIONS MANUAL FOR
Introduction to
Mathematical
Modeling and
Chaotic Dynamics

by

Ranjit Kumar Upadhyay

Satteluri R. K. Iyengar
SOLUTIONS MANUAL FOR
Introduction to
Mathematical
Modeling and
Chaotic Dynamics
by

Ranjit Kumar Upadhyay

Satteluri R. K. Iyengar

Boca Raton London New York

CRC Press is an imprint of the

Taylor & Francis Group, an informa business
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2014 by Taylor & Francis Group, LLC
CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Printed on acid-free paper

Version Date: 20130709

International Standard Book Number-13: 978-1-4822-2453-5 (Ancillary)

This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and
information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and
publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission
to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any
future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic,
mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or
retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact
the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides
licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment
has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation
without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Introduction to Mathematical Modeling

and

Chaotic Dynamics

Solution Manual

Dr. Ranjit Kumar Upadhyay

Dr. Satteluri R. K. Iyengar

1
 Chapter 1

Exercise 1.1

1. Let the dimensions of the box be: width = l, length = 2l, height = h. Volume = 2l 2 h = 20.
Cost of the total surface area of the open box is C = {30(l 3 + 10) / l}. Setting ( dC / dl ) = 0,
we get l 3 = 5. Substituting in C, we get the minimum cost as 90(52 / 3 )$. Since
( d 2C / dl 2 ) > 0 when l 3 = 5, the cost is a minimum.
2. Let x denote the number of Electronic Music Systems expected to be sold per week after
giving a rebate. The weekly increase in sale is ( x − 2000). The price function (demand
function) is given by p ( x ) = 3500 − {100( x − 2000) / 200} = 4500 − 0.5 x. The revenue
function is R( x ) = xp ( x ) = 4500 x − 0.5 x 2 . Setting R ′( x ) = 0, we get x = 4500. Since
R ′′( x ) < 0, the revenue is maximized. The corresponding price per unit should be
p ( x = 4500) = 2250. The rebate that should be offered is 3500 − 2250 = 1250$.
3. By Newton’s law of cooling, we obtain T − 20 = ( 90 − 20)e − kt = 70e − kt . The tempera-
ture is reduced to 600 C in 10 minutes. Hence, 4 = 7e −10k , or k = − ln( 4 / 7) / 10. To cool to
35 0 C , we require the total time t = [10 ln(15 / 70) / ln(4 / 7)] ≈ 27.5 minutes, that is another
17.5 minutes.
4. x1′ = −2 x 2 + r x1 , x2′ = 3x1 + r x 2 , r = ( x12 + x22 )1 / 2 .
Choose V ( x ) = 3x12 + 2 x22 . V * ( x ) = 2 rV . Unstable.
5. x1′ = −4 x1 + 8 x1 x22 , x2′ = −6 x2 − 12 x12 x2 .
Choose V ( x ) = [( 3x12 + 2 x22 ) / 24]. V * ( x ) = −( x12 + x22 ). Asymptotically stable.
6. x1′ = − x1 − x22 , x2′ = 2 x1 x2 − x23 .
Choose V ( x ) = ax12 + bx22 and obtain a = 2b. Let b = 1. V ( x ) = 2 x12 + x22 .
V * ( x ) = −2( 2 x12 + x22 ). Asymptotically stable.
7. x1′ = − x13 + x14 , x2′ = x14 − x23 .
Choose V ( x) = x12 + x22 . We get V * ( x ) = −2{ x14 + x 24 − x14 ( x1 + x 2 )}. In the
neighborhood of origin (0, 0), V * ( x ) < 0. Asymptotically stable.
8. x1′ = − x1 x22 p , x2′ = x12 q x2 , (p, q are positive integers).
Choose V ( x ) = x12 q + x22 p . V * ( x ) < 0, for p < q. Asymptotically stable in this case.
V * ( x ) = 0, for p = q. Stable.
9. x1′ = −4 x1 − x1 x22 , x2′ = −6 x2 − x12 x2 .
⎡− 2( 4 + x22 ) − 4 x1 x 2 ⎤
M ( x1 , x2 ) = ⎢ 2 ⎥
. det ( M 1 ) = −2( 4 + x22 ) < 0.
⎣ − 4 x1 x 2 − 2( 6 + x1 )⎦
det ( M 2 ) = 4( 24 + 4 x12 + 6 x22 − 3 x12 x22 ) > 0, in the neighborhood of origin. M 1 , M 2
are the leading minors of M. M is negative definite. Origin (0, 0) is asymptotically
stable.
2
10. x1′ = −2 x1 − 3x 2 − 5 x15 , x2′ = 3x1 − 2 x2 − 2 x25 .
⎡− 2( 2 + 25 x14 ) 0 ⎤
M ( x1 , x2 ) = ⎢ 4 ⎥
.
⎣ 0 − 4 (1 + 5 x 2 ⎦
)
M is negative definite. Origin (0, 0) is asymptotically stable.
11. z ′′′ + 9 z ′′ + 26 z ′ + 24 z = 0. Eigen values of A are −2, −3, −4. B = diag (1 / 4, 1 / 6, 1 / 8).
V ( y ) = (1 / 4) y12 + (1 / 6) y22 + (1 / 8) y 32 . Asymptotically stable.
12. z ′′′ − 4 z ′′ − 4 z ′ + 16 z = 0. Eigen values of A are −2, 2, 4. B = diag(1 / 4, − 1 / 4, − 1 / 8). V(y)
is not positive definite. Method cannot be applied.
13. P (λ ) = λ3 + 9λ2 + 6λ + 1. ai > 0 for all i and a1a 2 − a0 a 3 > 0. The roots of P(λ ) = 0 are
negative or have negative real parts.
14. P(λ ) = λ3 + 3λ2 + λ + 8. ai > 0 for all i and a1a 2 − a0 a3 < 0. One or more roots of
P(λ ) = 0 have positive real parts.
15. P (λ ) = ( 2 − 3 p )λ3 + ( 4 + p )λ2 + 2(1 + p )λ + p. For 0 < p < ( 2 / 3) , the roots of
P(λ ) = 0 are negative or have negative real parts.
16. x1′ = − x1 + 3e x 2 − 3 cos x1 , x 2′ = −1 + e x 2 − 6 x 2 − sin x1.
⎡− 1 3⎤
A=⎢ ⎥. Eigen values of A are −2, −4. A is stable. R(x) satisfies the condition
⎣− 1 − 5⎦
(1.16). By Theorem 1.5, zero solution is asymptotically stable.
17. x1′ = −1 − x2 + e x1 , x2′ = 4 x1 − 2 sin x2 .
⎡1 − 1⎤
A=⎢ ⎥. Eigenvalues of A are [{−1 ± i 7 } / 2] and have negative real parts. A is
⎣4 − 2 ⎦
stable. R(x) satisfies the condition (1.16). By Theorem 1.5, zero solution is asymptotically
stable.
18. x1′ = − x1 + 3e x 2 − 3 cos x1 , x 2′ = 1 + 6 x 2 − e x 2 − sin x1 .
⎡ 1 3⎤
A=⎢ ⎥. Eigenvalues of A are 2, 4. By Theorem 1.6, zero solution is unstable.
⎣− 1 5 ⎦
19. x1′ = − x2 − 4 x13 , x2′ = 9 x1 − 2 x23 .
⎡0 − 1⎤
A= ⎢ . Eigenvalues of A are λ = ±3i. Critical case. Choose
⎣9 0 ⎥⎦
V ( x1 , x2 ) = 9 x12 + x22 . V * = −4(18 x14 + x24 ). Zero solution is asymptotically stable.
20. The unique positive equilibrium point ( x1* , x2* ) is the solution of the equations
3 ln x1 + 2 ln x2 = 11, 4 ln x1 + 3 ln x2 = 13. We obtain x1* = e 7 , x2* = e −5 . Using the
transformation y1 = ln( x1 / x1* ) = ln x1 − 7, and y 2 = ln( x 2 / x 2* ) = ln x 2 + 5, we obtain the
linear system
⎡ y1′ ⎤ ⎡ − 3 − 2⎤ ⎡ y1 ⎤
⎢ y ′ ⎥ = ⎢− 4 − 3⎥ ⎢ y ⎥.
⎣ 2⎦ ⎣ ⎦ ⎣ 2⎦

3
 The characteristic equation associated with the above linear system is λ2 + 6λ + 1 = 0.
The roots of the characteristic equation have negative real parts. ( x1* , x2* ) is globally
asymptotically stable.
21. x1′ = − x2 + 6 x1 ( x12 + x22 − 3), x2′ = x1 + 6 x2 ( x12 + x22 − 3).
Period: 2π . C : x1 = 3 cos t , x2 = 3 sin t . Limit cycle is unstable.
22. x1′ = −3 x2 + 2 x1 (1 − x12 − x22 ), x 2′ = 3 x1 + 2 x 2 (1 − x12 − x 22 ).
Period: ( 2π / 3) . C : x1 = cos 3t , x2 = sin 3t. Limit cycle is Stable.
23. x1′ = − x2 + x1 (1 − x12 − x22 ) 2 , x2′ = x1 − x2 (1 − x12 − x22 ).
Period: 2π . C : x1 = cos t , x2 = sin t. I = 0. Study of the nonlinear terms is required to
decide whether C is semistable.
24. x1′ = −2 x1 + 3 x2 , x2′ = − x1 − 6 x23 . Choose V ( x) = x12 + 3 x 22 . Now, V ( x) > 0 and
V * ( x) < 0 for all ( x, y ) ≠ (0, 0). Hence, there are no closed orbits.
25. x1′ = −2 x1 + 4 x 23 − 4 x 24 , x2′ = − x1 − x2 + x1 x2 . Choose V ( x) = x12 + 2 x24 . Now,
V ( x) > 0 and V * ( x) < 0 for all ( x, y) ≠ (0, 0). Hence, there are no closed orbits.
26. x + μ ( x 4 − 1) x + x = 0, μ > 0 . We have f ( x ) = μ ( x 4 − 1), g ( x ) = x,
F ( x) = ( μ x / 5)( x 4 − 5), α = 5 . By Theorem 1.9 ( Lie ′nard' s Theorem), the system has a
unique limit cycle.
27. x + μ ( x 2 − 1) x + x 3 = 0, μ > 0. We have f ( x ) = μ ( x 2 − 1), g ( x ) = x 3 ,
F ( x) = ( μ x / 3)( x 2 − 3), α = 3 . By Theorem 1.9 ( Lie ′nard' s Theorem), the system has a
unique limit cycle.
28. Writing it as a system, we get x = y, y = −[ε ( x 6 − 1) y + x ]. Here, h ( x, y ) = ( x 6 − 1) y.
On the limit cycle, we have the approximation, x( t ) ≈ a cos t, y ( t ) ≈ −a sin t , T ≈ 2π . The
energy balance equation gives
2π
π
∫ [ a 6 (cos6 t − cos8 t ) + cos2 t − 1)]dt =
64
(5a 6 − 64) = 0.
0
The positive solution of this equation is a 0 = amplitude = 2 / 6 5. Also,
πε
g ′( a0 ) = − ( 40a 07 − 128a0 ) = −6πεa0 < 0.
64
The limit cycle is stable when ε > 0 and unstable when ε < 0.
29. x ′ = ( x / 2) − y , y ′ = x + ( y / 2 ).
⎛1 / 2 − 1⎞
A = ⎜⎜ ⎟ . The eigen values are λ = (1 / 2) ± i . We have p = 1 > 0,
⎝1 1 / 2 ⎟⎠
q = 5 / 4 > 0, Δ = p 2 − 4q = −4 < 0. Origin (0, 0) is an unstable focus.
30. x ′ = 3 x − 2 y , y ′ = 9 x − 3 y.
⎛3 − 2 ⎞
A = ⎜⎜ ⎟ . The eigen values are λ = ±3i . Here p = 0, q = 9 > 0. Therefore, (0, 0) is
⎝9 − 3 ⎟⎠
a centre and is stable.

4
31. x = −5 y − x ( x 2 + y 2 ) , r (0) = r0 , y = 5 x − y ( x 2 + y 2 ) , θ (0) = θ 0 . x 2 + y 2 ≠ 0.
Define f(0) = 0. We obtain r = −r 2 , and θ = 5. Origin is an equilibrium point of the
system. Integrating and using the initial conditions, we obtain
r = r0 /(1 + t r0 ), for t > 0. θ = 5t + θ 0 .
We find that r → 0 , and θ → ∞ as t → ∞ . Therefore, origin is a stable focus.
2y 2x
32. x = −3x − 2 2
, r (0) = r0 , y = −3 y + , θ ( 0) = θ 0 . x 2 + y 2 ≠ 0.
5 ln( x + y ) 5 ln( x 2 + y 2 )
Define f(0) = 0. We obtain r = −3r, r = r0e−3t ,
1 1 1 ⎡ 3t ⎤
θ = , θ = θ 0 − ln ⎢1 − .
15 ⎣ ln r0 ⎥⎦
=
5 ln r 5(ln r0 − 3t )
For, r0 < 1, r( t ) → 0 and θ → ∞ , as t → ∞. Origin is a stable focus in this case.
33. van der Pol oscillator x − ε (1 − x 2 ) x + x = 0, where ε > 0, is a damping constant.
div( F ) = ε (1 − x 2 ) < 0 for x > 1. System is dissipative for x > 1. div( F ) > 0 for
x < 1. The system is an area expanding dynamical system for x < 1.
34. Lorenz system (a simplified model of atmospheric convection).
dx dy dz
= σ ( y − x ); = − xz + rx − y; = xy − bz. b and σ are positive constants.
dt dt dt
div( F ) = −(σ + 1 + b) < 0. System is dissipative.
35. Rossler equation : x = −( y + z ), y = x + ay, z = b + xz − cz, with a = b = 0.2 and c = 5.7.
div( F ) = a + x − c = −5.5 + x < 0 for 0 < x < 5.5. System is dissipative for 0 < x < 5.5.
Area expanding for x > 5.5.
36. The Jacobian matrix of the system is given by
⎡ ∂x n + 1 ∂x n + 1 ⎤
⎢ ∂x ∂y n ⎥ ⎡1 − a − 2 yn ⎤
J(X ) = ⎢ n ⎥=⎢ ⎥.
⎢ ∂y n +1 ∂y n +1 ⎥ ⎣ y n 1 + b + x n ⎦
⎢⎣ ∂x n ∂y n ⎥⎦
Now, det( J ( X ) = (1 − a )(1 + b + x n ) + 2 y n2 > 1 for a ≤ 0 and b ≥ 0.
Therefore, Burgers map is area expanding for these cases.
37. The Jacobian matrix of the system is given by
⎡ ∂xn +1 ∂xn +1 ⎤
⎢ ∂x ∂yn ⎥ ⎡− 2( a + c) xn 0 ⎤
J(X ) = ⎢ n ⎥=⎢ .
⎢ ∂y n +1 ∂y n +1 ⎥ ⎣ 2b − 2( a − c ) y n − 2b⎥⎦
⎢⎣ ∂xn ∂yn ⎥⎦
Now, det( J ( X ) = 4( a 2 − c 2 ) x n y n + 4b( a + c ) xn . For a = c > 0, b > 0,
det( J ( X ) = 8abxn < 8ab, in the unit square, 0 < xn < 1, 0 < yn < 1 . Therefore, the map
is dissipative when 8ab < 1 , and area expanding when 8ab > 1.

38. x = y − x , y = 6 + y + x (4 − 2 x − x 2 ). Hamiltonian system.

5
 H ( x, y ) = −6 x − xy − 2 x 2 + ( 2 / 3) x 3 + (1 / 4) x 4 . Equilibrium points are (−1, −1), (2, 2),
(−3, −3). All points are saddle points.
39. x = y ( 4 − x 2 ), y = x ( x 2 + y 2 + 5). Hamiltonian system.
H ( x, y ) = (8 y 2 − 2 x 2 y 2 − 10 x 2 − x 4 ) / 4. Equilibrium point (0, 0) is a saddle point.
40. x = 4 − x 2 − y 2 , y = x( x 2 + y 2 + 3). Not a Hamiltonian system.
41. x = y ( 4 + 5 x + x 2 ), y = [ x(1 − 2 y 2 ) − 5 y 2 ] / 2. Hamiltonian system.
H ( x, y ) = [(8 + 10 x + 2 x 2 ) y 2 − x 2 ] / 4. Equilibrium points are (0, 0), ( −4, ± 2 / 3 ) . All
are saddle points.
42. x = y ( 4 − 5 x − x 2 ), y = [(5 + 2 x ) y 2 − x ] / 2. Hamiltonian system.
H ( x, y ) = [(8 − 10 x − 2 x 2 ) y 2 + x 2 ] / 4. Equilibrium points are (0, 0), ( x0 , ± y0 ), where
x0 = ( −5 + 41) / 2, y0 = ( 41 − 5 41) / 82 ) . (0, 0) is a centre. The remaining two points
are saddle points.
43. Lotka-Volterra population model governing the growth/ decay of species
x = ax − bxy, y = cxy − dy, for a = 1.0 , b = 0.2, c =0.4, d =0.5.
% Matlab code for Lotka-Volterra model in Problem 43

g=inline('[1.*x(1)- (0.2.*x(1).*x(2));0.5.*x(2)+(0.04.*x(1).*x(2))]','t','x');
[t xa]=ode45(g,[0 150],[10 10])
plot(xa(:,1),xa(:,2))
%plot3(xa(:,1),xa(:,2),xa(:,3));
%comet3(xa(:,1),xa(:,2),xa(:,3));
hold on;
figure;
plot(t,xa(:,1),'r');
hold on;
plot(t,xa(:,2),'g');
The phase plot and Time series are given in Figs. 1.1 and 1.2.
11
x
35 y
10

9 30

8
25
7
Population density

20
6
y

5 15

4
10
3
5
2

1 0
0 5 10 15 20 25 30 35 0 50 100 150
x Time

Fig.1.1. Phase plot for Problem 43. Fig.1.2. Time series for Problem 43.

44. R o ssler model x = −( y + x ), y = x + ay, z = b + z ( x − c), when a = 0.2, b = 0.2 , c =

5.7 .
% Matlab code for the R o ssler model
g=inline('[-(x(2)+x(3));x(1)+0.2.*x(2);0.2+x(3).*(x(1)-5.7)]','t','x');
[t xa]=ode45(g,[0 500],[0.5 0.5 0.5])
6
 plot3(xa(:,1),xa(:,2),xa(:,3))
hold on;
figure;
plot(t,xa(:,1),'r');
hold on;
plot(t,xa(:,2),'g');
plot(t,xa(:,3),'b');
The phase plot and Time series are given in Figs. 1.3 and 1.4.
Chaotic Attractor
x
25 y
z
20
25

20 15

15 10

State Variables
z

10 5

5 0

0
-5
10
15
0 10 -10
5
-10 0
-5 -15
y -20 -10 400 450 500 550 600 650
x
Time

Fig.1.3. Phase plot for Problem 44. Fig.1.4. Time series for Problem 44.

45. Phase plot and the corresponding time series using MATLAB for the given model.
when a1 = 2.5, b1 = 0.05, w = 0.85, α = 0.45, β = 0.2, γ = 0.6, a 2 = 0.95, and w1 = 1.65.
% Matlab code for Problem 45
g= inline('[2.5.*u(1)−0.05.*u(1).*u(1) − (0.85.*u(1).*u(2))./(0.45+0.2.*u(2)+0.6.*u(1));
−0.95.*u(2)+(1.65.*u(1).*u(2))./(0.45+0.2.*u(2)+0.6.*u(1))]','t','u');
[t ua]=ode45(g,[400 550],[5.9157 35.6279])
figure;
plot(t,ua(:,1),'b');
hold on;
plot(t,ua(:,2),'r');
xlabel('time');
ylabel('state variable');
legend('x','y');
figure;
plot(ua(:,1),ua(:,2));
xlabel('x');
ylabel('y');
The phase plot and Time series are given in Figs. 1.5 and 1.6.
40

35

30

25

20
y

15

10

0
0 2 4 6 8 10 12 14 16 18 20
x

Fig.1.5. Phase plot for Problem 45. Fig.1.6. Time series for Problem 45.

7
46. Basin boundary for the limit cycle and chaos for the food-chain model (Rai and
Upadhyay [24]) when a1 = 1.75, b1 = 0.05, a 2 = 1.0, c = 0.7, w = 1.0, w1 = 2.0, w2 = 1.5,
w3 = 3.75, D = 10.0, D1 = 10.0, D2 = 10.0, D3 = 20.
Basin boundaries for the chaotic attractor are plotted in Fig. 1.7.

Fig.1.7. Basin boundaries for the chaotic attractor for Problem 46.
(From Rai, V., Upadhyay, R. K., Chaotic population dynamics and biology of the top-predator.
Chaos, Solitons Fractals, 21, 1195–1204, Copyright 2004, Elsevier. Reprinted with permission).

CHAPTER 2
Exercise 2.1

1. Without loss of generality, let the initial time be t = 0. The solution is given by
P(t ) = P0 e rt . At time t = 0, we have P = P0 . Let at time t = T , the population is tripled,
that is P(T ) = 3P0 . Hence, 3P0 = P0 e rT , or T = ln 3 / r. The time taken for the population
to triple its size is T = ln 3 / r.
2. We have r = 0.09, K = 900, t0 = 0, P0 = 90. The value of the constant A is obtained as
A = [( P0 − K ) / P0 ] = −9. The solution is given by
K 900 900
P (t ) = − rt
= − 0.09t
. P(90) = ≈ 897.
1 − Ae 1 + 9e 1 + 9e − 8.1
3. (i) Equilibrium solutions are P1 = 0 and P2 = 500. Now,
⎛ P ⎞
f ′( P ) = 0.5⎜1 − ⎟, f ′( P1 ) = 0.5 > 0, f ′( P2 ) = −0.5 < 0.
⎝ 250 ⎠
Hence, the equilibrium point P1 = 0 is unstable and the equilibrium point P2 = 500 is
stable. Thus, solutions initiating in a neighbourhood of K = 500 approach K as t → ∞
while no solution starting in a neighbourhood of P = 0 remains close to zero in the future.
(ii) P (t ) = 500 /[1 + 9e −0.5t ].
dP ⎛ P ⎞ K 100
4. = (0.0025) P⎜1 − ⎟. The solution is P(t ) = −
= ,
dt ⎝ 100 ⎠ 1 − Ae rt
1 − Ae − 0.0025t
where A = −11.5 e 4.9875. The estimated populations in the years 2005 and 2100 are 8.18
and 10.15 billions respectively.

8
5. Separating the variables and solving the differential equation, we get R /( R − 1) = ceαt , a,
c are arbitrary constants to be determined. Applying the conditions that at 8 hours, R =
0.08, and at 12 hours R = 0.05, we obtain a = 0.25 ln(11.5), and c = − e −12a ≈ − 1 / 1520.9.
Now, when R = 0.9, we obtain t = 15.5985 or approximately 3.36 P.M.
α ⎤ ⎡ ⎛ P ⎞α ⎤
d 2P ⎛ r ⎞ ⎡ ⎛ P ⎞
2
6. We have = ⎜ ⎟ ⎢ ⎜ ⎟
P 1 − (1 + α ) ⎥ ⎢1 − ⎜ ⎟ ⎥
dt 2 ⎝ α ⎠ ⎢⎣ ⎝ K ⎠ ⎥⎦ ⎢⎣ ⎝ K ⎠ ⎥⎦
2 α
⎛r⎞ ⎛P⎞
= ⎜ ⎟ P[1 − t ( 2 + α ) + t 2 (1 + α )], where t = ⎜ ⎟ .
⎝α ⎠ ⎝K⎠
Setting P ′′( t ) = 0, we get P = 0, t = 1, and t = 1 /(1 + α ), that is, P = 0, P = K, and P =
K (1 + α ) −1 / α . The point of inflection is P = K (1 + α ) −1 / α .
For α = 1, the point of inflection is P = K/2. Taking the limit as α → 0, we find that the
point of inflection moves to P = K/e.
7. Comparing with (2.12), we have r = 10 −3 , γ = 10 −9 , P0 = 105. The solution is given by
rP0 106
P (t ) = = .
γ P0 + ( r − γ P0 ) e − rt 1 + 9 exp( −0.001t )
The limiting value of the population is 106.
8. Comparing with the model for the density dependent growth (2.12), we have r = 0.09,
γ = 0.0009, P0 = 2500. The solution of the model is obtained as
rP0
P (t ) =
γ P0 + ( r − γ P0 )e − rt
(0.09)( 2500) 225
= − 0.09t
= .
(0.0009)( 2500) + {0.09 − 0.0009( 2500)}e 2.25 − 2.16e − 0.09t
The limiting value of the population is 100. The equilibrium points are 0 and 100. The
point 0 is unstable and the point 100 is stable. All other solutions move away from P = 0,
towards P =100.
9. P(t ) = P0 exp[( r0 / α )(1 − e −αt )].
10. Assume that no student leaves the campus throughout the duration of viral fever. Now,
Number of students infected at a point of time = p,
Number of students who are not infected at the same point of time = 2500 − p.
Since, the rate at which viral fever spreads is proportional not only to the number of
people affected, but also to the number of people who are not yet exposed to it, we obtain
the model as
dp
= ap( 2500 − p ), p(0) = 1, a is an arbitrary constant.
dt
Separating the variables and solving, we get p (t ) = 2500 /(1 + 2499e −2500at ). Since,
p (5) = 50, we obtain a = ln(51) / 12500. Hence, p (15) ≈ 2454 students.
11. The equilibrium points are the solutions of the equation P(6 − P ) − h = 0. The
equilibrium points are given by P1 = 3 − 9 − h , P2 = 3 + 9 − h , h < 9. We have
f ′( P ) = 6 − 2 P, f ′( P1 ) = 2 9 − h > 0, f ′( P2 ) = −2 9 − h < 0.
9
 The equilibrium point P1 is unstable whereas the equilibrium point P2 is asymptotically
stable. The two equilibrium points coincide when h = 9. The critical value of h is hc = 9.
When h > 9 , there are no equilibrium points and the population tends to extinction. If
P > 3, f ′( P) < 0. In this case, we get stable solutions. Hence, the initial value of the
population should satisfy P0 > 3. That is, the population may be extinct if P0 < 3 , even
though the condition h < hc is satisfied.
12. The equation is a harvesting model with constant harvesting.
(i) The equilibrium points are
1 K 4h
N = [ rK ± r 2 K 2 − 4 rhK ] = [1 ± 1 − p ], where p = .
2r 2 rK
K K
Let, N 1 = [1 − 1 − p ], and N 2 = [1 + 1 − p ].
2 2
Real solutions are obtained only when p < 1, or h < (rK / 4). We find
f ′( N ) = r[1 − (2 / K ) N ], f ′( N 1 ) = r 1 − p > 0, f ′( N 2 ) = − r 1 − p < 0.
The equilibrium point N 1 is unstable whereas the equilibrium point N 2 is asymptotically
stable. The two equilibrium points coincide when p = 1, that is when 4h = rK , or
h = rK / 4 and N = K / 2. The critical value of h is hc = rK / 4. When h > rK / 4 , there
are no equilibrium points and the population tends to extinction.
(ii) If 2 N > K , f ′( N ) < 0. In this case, we get stable solutions. Hence, the initial value
of the population should satisfy n0 > ( K / 2). That is, the population may become extinct
if n0 < ( K / 2 ) , even though the condition h < hc is satisfied.
13. Equilibrium point is N 1 = Ke − qE / α . N = 0 can not be taken as an equilibrium point even
though the limit of the right hand side exists. Now,
f ′( N ) = α [ln( K / N ) − 1] − qE , and f ′( N 1 ) = −α < 0.
Therefore, N 1 is asymptotically stable. Note that f ′( N = 0) is not defined.
The sustainable yield is given by
Y ( qE ) = qEN1 = qEKe − qE / α = α Kue − u = Y (u ), where u = qE / α .
Also, Y ′(u ) = αK [1 − u ]e − u , Y ′′(u ) = αK [u − 2]e − u . Setting Y ′(u ) = 0, we get the
stationary point as u = 1. When u = 1, Y ′′(u ) < 0. The maximum occurs for u = 1, or
qE = α . Maximum sustainable yield = Y (u = 1) = αK / e.
14. The location of the steady state varies with the length of the delay and the form of the
characteristic equation changes due to the direct inclusion of the delay in the parameters
and the indirect changes resulting from the varying location of the steady state. The
model has the trivial steady state (0, 0) and the nontrivial steady state is given by
1 ⎛ be − μτ ⎞⎟
be − μτ e − aP = d , or P = ln⎜ .
a ⎜⎝ d ⎟⎠
In particular, if τ > (1 / μ ) ln(b / d ), there is no positive steady state. In this case, given
that the initial value is positive, we have
dP
≤ be − μτ P (t − τ ) − dP(t ), with be − μτ < d .
dt
The solution goes to zero and the trivial steady state is globally stable.
10
15. P* = F ( P*), gives P* = K (1 + r ) which depends on r. At P = P*,
F ′ = [1 /(1 + r )] < 1 for all r > 0. Asymptotically stable for all r.
16. We find the solution of the equation P* = F ( P*), that is of
⎡ ⎛ P * ⎞⎤ ⎡ ⎛ P * ⎞⎤
P* = P * exp ⎢λ ⎜1 − ⎟⎥, or exp ⎢λ ⎜1 − ⎟ = 1.
⎣ ⎝ K ⎠⎦ ⎣ ⎝ K ⎠⎥⎦
The solution is given by P* = K, for all λ .
⎡ ⎛ P ⎞⎤ dF ⎡ ⎛ P ⎞⎤ ⎡ λ P ⎤
Now, F ( P ) = P exp ⎢λ ⎜1 − ⎟⎥, = exp ⎢λ ⎜1 − ⎟⎥ ⎢1 − .
⎣ ⎝ K ⎠⎦ dP ⎣ ⎝ K ⎠⎦ ⎣ K ⎥⎦
dF
At P = P* = K , we get = [1 − λ ].
dP
Equilibrium is stable for 1 − λ < 1, or 0 < λ < 2. It is unstable for λ < 0 and λ > 2.
Neighboring trajectories approach the equilibrium point asymptotically for 0 < λ < 1, and
with damped oscillations for 1 < λ < 2.
dF K d 2F ⎛ r ⎞ ⎡ ⎛ P ⎞⎤ ⎡ rP ⎤
Now, = 0, when P = . = ⎜ − ⎟ exp ⎢ r⎜1 − ⎟⎥ ⎢2 − ⎥.
dP r dP 2
⎝ K⎠ ⎣ ⎝ K ⎠⎦ ⎣ K⎦
K d 2F ⎛ r ⎞
For P = , = ⎜ − ⎟ exp( r − 1) < 0.
r dP 2 ⎝ K ⎠
Maximum of the trajectory occurs at P = K / r. The maximum value = ( K / r ) exp( r − 1).
(i) For K = 500, r = 1, P0 = 50, we have the following sequence of values: 122.98, 261.40,
421.26, 493.11, 499.95, 500, (see Fig.2.1).
(ii) For K = 500, r = 1, P0 = 660, we get the sequence of values: 479.26, 499.56, 500, (see
Fig. 2.1).

Fig.2.1. Discrete solution values for the data sets (i) and (ii).

17. Choose Vi = ( N i − K ) 2 . Vi ≥ 0 and has a minimum V = 0 at N i = K .

The increment ΔVi on the trajectory is given by
ΔVi = K 2 ni [ e r (1− ni ) − 1][ ni e r (1− ni ) + ni − 2], where N i = Kni .
ΔVi ≤ 0 when (i) e r (1− n i ) < 1, and [e r (1− ni ) + 1] ≥ [2 / ni ];
or (ii) e r (1− ni ) ≥ 1, and [e r (1− ni ) + 1] < [2 / ni ]. We obtain ΔVi ≤ 0 for 0 < r < 2 and for all
n i . ΔVi = 0 only for ni = N * / K . Hence, equilibrium is globally asymptotically stable.
11
18. Equilibrium point is x* = 1. F ′(1) = 1 − r < 1 gives 0 < r < 2. Asymptotically stable
for 0 < r < 2. Maximum occurs at x = 1/r and the maximum value is exp( r − 1) / r .
19. For steady state, we solve P* = F ( P*) = [( λP*) /(1 + bP *2 )] , to get P* = 0, and
P* = (λ − 1) / b , λ > 1. We have
dF λ (1 − bPn2 )
= .
dP (1 + bPn2 ) 2
Hence, the eigen values corresponding to P* = 0 and P* = (λ − 1) / b , are λ and
( 2 − λ ) / λ respectively. For λ =1, we have only a trivial equilibrium point.
Setting dF / dP = 0, we get P = 1 / b . For this value of P, F ′′( P ) < 0. Hence, we obtain
the maximum at P = 1 / b . The maximum value is F (1 / b ) = λ /(2 b ).

Exercise 2.2

1. Setting F ( X , Y ) = 0, G ( X , Y ) = 0, we obtain the equilibrium point as X * = 212 / 3,

Y * = 2332 / 15. Note that (0, 0) is not an equilibrium point. The elements of the Jacobian
matrix of the system evaluated at an equilibrium point (X*, Y*) are
X 8.0Y 2 1.6 X 2
a11 = 1 − − , a12 = − ,
50 ( X + 5Y ) 2 ( X + 5Y ) 2
3Y 2 0.6 X 2
a21 = , a 22 = − 0.05.
( X + 5Y ) 2 ( X + 5Y ) 2
At the equilibrium point ( X *, Y *) : We obtain a11 = −0.682222, a12 = −0.011111,
a 21 = 0.100833, a 22 = −0.045833. The characteristic equation is λ2 + 0.72806λ +
0.032389 = 0. The coefficients in the equation are positive. By Routh-Hurwitz criterion,
the eigen values are negative or have negative real parts. The equilibrium point
( X *, Y *) is asymptotically stable.
2. We have F ( P, Z ) = ( a1 − b1 P − c1Z ), and G ( P, Z ) = [a 2 − c2 ( Z / P )].
Conditions (i), (ii), (iii), (v), (vi) and (vii) are satisfied. Equality condition in (iv) is
satisfied. The requirement (viii), G (C, 0) = 0 gives a 2 = 0, which violates the
assumption that a 2 is positive. Hence, Kolmogorov theorem cannot be applied.
wY wX
3. We have F ( X , Y ) = a1 − b1 X − , G ( X , Y ) = −a2 + 1 .
X +D X + D1
Conditions (i), (ii), (iv), (v), (vi), (vii) are satisfied. Equality condition in (iii) is satisfied.
wC a D
(viii) G (C , 0) = −a2 + 1 = 0, gives C = 2 1 .
C + D1 w1 − a 2
C > 0 gives the condition w1 > a 2 .
a D1a 2 Da
(ix) B > C gives the condition 1 > , or w1 − a 2 > 1 2 .
b1 ( w1 − a 2 ) K
The two species system (2.80), (2.81) qualifies as a Kolmogorov system when the
conditions w1 > a2 , K ( w1 − a 2 ) > D1a 2 are satisfied.

12
4. The elements of the Jacobian matrix of the system are
wDY wX
a11 = a1 − 2b1 X − 2
, a12 = − ,
( X + D) X +D
w1 D1Y w1 X
a 21 = , a 22 = − a 2 + .
( X + D1 ) 2 X + D1
At the equilibrium point E0 (0, 0) : We obtain a11 = a1 , a12 = 0, a 21 = 0, a22 = −a 2 . The
eigen values are λ1 = a1 > 0, and λ2 = −a 2 < 0. The equilibrium point is unstable. Since
Re(λ ) ≠ 0 for both eigen values, the fixed point is hyperbolic. Since the eigen values are
real and are of opposite signs, we find that E0 is a hyperbolic saddle point which repels in
the x-direction and attracts in y-direction.
At the equilibrium point E1 ( K , 0) : We obtain
wK wK a
a11 = a1 − 2b1 K , a12 = − , a 21 = 0, a 22 = −a 2 + 1 , K = 1.
( K + D) K + D1 b1
The eigen values are λ1 = a11 = −a1 < 0, and
w1a1 a ( w − a 2 ) − a 2 b1D1
λ2 = a 22 = −a 2 + = 1 1 >0
b1 ( K + D1 ) a1 + b1 D1
using the result from Kolmogorov condition (ix). The equilibrium point is unstable. The
fixed point E1 ( K , 0) is also a hyperbolic saddle point which attracts in the x-direction and
repels in y-direction.
At the equilibrium point E*(X*, Y*): We obtain
wDY * wX *
a11 = a1 − 2b1 X * − 2
, a12 = − ,
( X * +D) X * +D
w1 D1Y * wX
a 21 = 2
, a 22 = G ( X , Y ) = −a 2 + 1 = 0.
( X * + D1 ) X + D1
The eigen values of J are the roots of λ2 − a11λ − a12 a 21 = 0. By Routh-Hurwitz theorem,
the necessary and sufficient conditions for the eigen values to be negative or have negative
real parts are ( −a11 ) > 0, ( −a12 a 21 ) > 0 . That is
⎛ wX * ⎞⎛⎜ w1 D1Y * ⎞⎟
( −a12 a 21 ) = ⎜ ⎟ > 0, which is true.
⎝ X * + D ⎠⎜⎝ ( X * + D1 ) 2 ⎟⎠
⎡ wDY * ⎤ wY * D
( −a11 ) = − ⎢a1 − 2b1 X * − 2⎥
= . − ( a1 − 2b1 X *)
⎣ ( X * + D) ⎦ ( X * + D) ( X * + D)
D
= ( a1 − b1 X *) − (a1 − 2b1 X *) (from F ( X , Y ) = 0 )
( X * +D)
X* ⎡ X * −K ⎤
= −(a1 − b1 X *) + b1 X * = b1 X * ⎢1 + ⎥
( X * + D) ⎣ ( X * + D) ⎦
⎡2X * +D − K ⎤
= b1 X * ⎢ ⎥.
⎣ ( X * + D) ⎦
Now, ( −a11 ) > 0, if 2 X * + D − K > 0. Substituting the expression for X*, we get

13
 ⎛ a D ⎞ ⎛ a D ⎞
2⎜⎜ 2 1 ⎟⎟ + D − K > 0, or 2b1 ⎜⎜ 2 1 ⎟⎟ + b1 D − a1 > 0. (A)
⎝ w1 − a2 ⎠ ⎝ w1 − a 2 ⎠
The equilibrium point E * ( X *, Y *) is locally asymptotically stable if the condition (A) is
satisfied.
5. The elements of the Jacobian matrix of the system are
hP v u
a11 = 1 − 2u − , a12 = − ,
( hP + u ) 2 hP + u
bhP v bu 2 fvh Z2
a 21 = , a 22 = −c− 2 .
( hP + u ) 2 (h P + u ) (hZ + v 2 ) 2
At the equilibrium point E0 (0, 0) : We obtain a11 = 1, a12 = 0, a 21 = 0, a 22 = −c. The eigen
values are λ1 = 1, and λ2 = −c < 0. The equilibrium point is a hyperbolic saddle point.
At the equilibrium point E1 (1, 0) : We obtain
1 b
a11 = −1, a12 = − , a 21 = 0, a 22 = − c.
(1 + hP ) hP + 1
b (b − c) − ch p
The eigen values are λ1 = a11 = −1 < 0, and λ2 = a 22 = −c = .
hP + 1 hP + 1
But (from text), b > c, 0 < chP < b − c, 0 < hP < 1. Therefore, the equilibrium E1 is a
saddle point with stable manifold locally in the u -direction and with unstable manifold
locally in the v -direction.
At the equilibrium point E * (u*, v*) : We have
hP v hP v
a11 = 1 − 2u − 2
=1− u − u − ,
(hP + u ) ( hP + u ) 2
v hP v ⎡ v ⎤
= −u− = u ⎢ − 1⎥.
( hP + u ) ( hP + u ) 2 ⎢⎣ ( hP + u )
2
⎥⎦
u bhP v bu 2 fvhZ2 fv( v 2 − hZ2 )
a12 = − , a 21 = , a 22 = − c − = ,
hP + u (hP + u ) 2 (h P + u ) (hZ2 + v 2 ) 2 ( hZ2 + v 2 ) 2
where all the quantities are evaluated at ( u*, v*). The eigen values of J are the roots of
λ 2 − (a11 + a22 )λ + (a11a22 − a12 a21 ) = 0. Using the Routh-Hurwitz theorem, we find the
conditions for local stability as
⎡ v ⎤ fv(hZ2 − v 2 )
A = −( a11 + a 22 ) = u ⎢1 − 2⎥
+ 2 2 2
> 0,
⎣⎢ ( hP + u ) ⎦⎥ ( hZ + v )
⎡ v ⎤ ⎡ fv( hZ2 − v 2 ) ⎤ bhP uv
B = a11a 22 − a12 a 21 = u ⎢1 − 2 ⎥⎢ 2 2 2 ⎥
+ 3
> 0,
⎢⎣ ( hP + u ) ⎥⎦ ⎢⎣ ( hZ + v ) ⎥⎦ (h P + u )
where all the quantities are evaluated at ( u*, v*). Sufficient conditions are
v* < ( hP + u*) 2 , and ( v*) 2 < hZ2 . If the above conditions are satisfied, then both predator
and prey species coexist, and they settle down at its equilibrium point.

14
6. The non-zero equilibrium points are the solution of the equations
wY w1 X
a1 − b1 X − = 0, − a2 + = 0.
(α + βY + γX ) (α + βY + γX )
a1 − b1 X 1 a
We have = = 2 . Solving the right equality, we obtain
wY α + βY + γX w1 X
1
Y = [( w1 − γa 2 ) X − αa 2 ]. (2.1)
a2 β
Using the first and third terms in the equality and substituting the expression for Y, we
w
obtain ( a1 − b1 X ) w1 X = [( w1 − γa 2 ) X − αa 2 ].
β
Simplifying, we obtain β w1 b1 X 2
+ [ w ( w1 − γ a 2 ) − β w1 a 1 ] X − w α a 2 = 0 .
The roots of this equation are X = [ p ± p 2 + q ] /(2 βw1b1 ),
where p = βw1a1 + wγa 2 − ww1 , q = 4 βw1b1wαa 2 . Irrespective of the sign of p, the root
in the first quadrant is X * = [ p + p 2 + q ] /(2 βw1b1 ).
The value of Y* is given by (2.1) (the chosen parameter values should satisfy Y* > 0).
wY w1 X
7. We have F ( X , Y ) = a1 − b1 X − , and G ( X , Y ) = −a2 + .
(α + βY + γX ) (α + β Y + γ X )
Conditions (i), (ii), (iii), (iv), (v), (vii) are satisfied.
a1α
(vi) F (0, A) = 0, gives A= > 0 if w > a1 β .
( w − a1 β )
a 2α
(viii) G (C , 0) = 0, gives C = > 0 if w1 > a 2γ .
( w1 − a2γ )
a a2α
(ix) B > C gives 1 > .
b1 ( w1 − a2 γ )
a a2α
Summarizing, we get the conditions as w > a1 β , w1 > a2 γ and 1 > .
b1 ( w1 − a2 γ )
An oscillatory predator-prey dynamics (time series) exhibited by the model system for
the given set of parameter values, a1 = 2.5, b1 = 0.05, w = 0.85, α = 0.45, β = 0.2,
γ = 0.6, a 2 = 0.95, and w1 = 1.65, is presented in Fig. 2.2.
8. The equilibrium points are (0, 0), (a1 / b1 , 0) and (X*, Y*) (see Problem 6).
The elements of the Jacobian matrix of the system are
w(α + β Y )Y w(α + γ X ) X
a11 = a1 − 2b1 X − , a12 = − ,
(α + β Y + γ X ) 2
(α + β Y + γ X ) 2
w1 (α + β Y )Y w (α + γ X ) X
a21 = , a22 = −a2 + 1 .
(α + β Y + γ X ) 2
(α + β Y + γ X )2

15
 Fig.2.2. Time-series displaying oscillatory dynamics in the model (2.71)-(2.72).

At the equilibrium point ( 0, 0) : We obtain a11 = a1 , a12 = 0, a21 = 0, a22 = −a2 . The eigen
values are λ1 = a1 , and λ2 = −a2 < 0. The equilibrium point (0, 0) is unstable. Since,
Re(λ ) ≠ 0 for both eigen values, the fixed point is hyperbolic. Since, the eigen values are
real and are of opposite signs, we find that (0, 0) is a hyperbolic saddle point which repels
in the x-direction and attracts in y-direction.
At the equilibrium point (K, 0), where K = a1 / b1.
wK wK
We obtain a11 = −a1 , a12 = − , a 21 = 0, a 22 = 1 − a2 .
α + γK α + γK
The eigen values are λ1 = − a1 , and λ2 = a22 = ( w1 K /(α + γ K )) − a2 . If λ2 < 0, that is
[ w1K /(α + Kγ )] < a 2 , that is, [ w1a1 /(b1α + a1γ )] < a2 , then the equilibrium point (K , 0) is
asymptotically stable. Otherwise, (K , 0) is unstable. It depends on the values of the
parameters w1 , a1 , a 2 , b1 , α , γ . If λ2 > 0, that is, [ w1a1 /(b1α + a1γ )] > a 2 , then the eigen
values are real and are of opposite signs and the fixed point (K , 0) is a hyperbolic saddle
point.
At the equilibrium point (X*, Y*): The expressions for X*, Y* are
X* = [p + p 2 + q ] /(2 βw1b1 ); p = βw1a1 + wγa 2 − ww1 , q = 4 βw1b1wαa 2 .
a 2 β Y * = [( w1 − γa 2 ) X * −αa2 ].
The eigen values of J are the roots of λ2 − ( a11 + a 22 )λ + ( a11a 22 − a12 a 21 ) = 0. Using the
Routh-Hurwitz theorem, the necessary and sufficient conditions are given by
− ( a11 + a 22 ) > 0, and ( a11a 22 − a12 a 21 ) > 0. We have (dropping *)
a1 − b1 X 1 a
= = 2 . (see Problem 6)
wY α + βY + γX w1 X
1 1
α + βY = α + [( w1 − γa 2 ) X − αa 2 ] = ( w1 − γa 2 ) X ..
a2 a2
w(α + βY )Y
a11 = a1 − 2b1 X −
(α + βY + γX )2
⎡ w( w1 − γa 2 ) X ⎤ ⎡ ( a1 − b1 X )a 2 ⎤ γa 2
= a1 − b1 X − b1 X − ⎢ ⎥ ⎢ ww X ⎥ = w ( a1 − b1 X ) − b1 X .
⎣ a2 ⎦⎣ 1 ⎦ 1
w (α + γX ) X a (α + γX ) a β
a 22 = −a2 + 1 = −a2 + 2 = − 2 ( a1 − b1 X ).
(α + βY + γX ) 2 (α + βY + γX ) w
16
 γ a2 a β
a11 + a 22 = ( a1 − b1 X ) − b1 X − 2 ( a1 − b1 X )
w1 w
⎡γ β⎤ ⎡ γa βa ⎤
= a1a 2 ⎢ − ⎥ − b1 X ⎢ 2 + 1 − 2 ⎥
⎣ w1 w ⎦ ⎣ w1 w ⎦
⎡ γw − β w1 ⎤ b1 X
= a1a 2 ⎢ ⎥+ [a2 ( βw1 − γw) − ww1 ].
⎣ ww1 ⎦ ww1
Sufficient conditions for [ −( a11 + a 22 )] > 0 are
γw − βw1 < 0, and a 2 ( βw1 − γw) − ww1 < 0. Now,
⎡ γa ⎤⎡ a β ⎤
( a11a 22 − a12 a 21 ) = ⎢ 2 (a1 − b1 X ) − b1 X ⎥ ⎢ − 2 ( a1 − b1 X )⎥
⎣ 1w ⎦⎣ w ⎦
wa 22
+ (α + γX )( w1 − γa2 )( a1 − b1 X )
w13 X
a β
= 2 ( a1 − b1 X )[b1 X ( w1 + γa 2 ) − γa1a2 ] + second term.
ww1
Sufficient conditions for ( a11a 22 − a12 a 21 ) > 0 are
( w1 − γa2 ) > 0, ( a1 − b1 X ) > 0, and b1 X ( w1 + γa2 ) − γa1a2 > 0,
γa1a2
that is, ( w1 − γa 2 ) > 0, and < b1 X < a1 .
( w1 + γa 2 )
We require the above conditions to be satisfied for asymptotic stability. However, it is
possible to derive alternate conditions by simplifying in a different way.

9. The equilibrium points are the solutions of the equations

⎡⎛ u⎞ v ⎤ ⎡ βu ⎤
u ⎢⎜ 1 − ⎟ − 2 ⎥ = 0, v ⎢ 2 −γ ⎥ = 0.
⎣⎝ K ⎠ ( u / α ) + u + 1 ⎦ ⎣ (u / α ) + u + 1 ⎦
Two of the equilibrium points are (0, 0) and (K, 0). From the second equation, we get
[1 /{(u 2 / α ) + u + 1}] = (γ / β u ). Using this result in the first equation, we get
v* = ( β / γK )( K − u*)u * .
Simplifying the equations ( K − u )( u 2 + α u + α ) − vKα = 0, αβ u − γ (u 2 + α u + α ) = 0 ,
we obtain ( K − u ) [u 2 + α u + α − (αβ / γ )u ] = 0. The first root gives u = K, which gives
the equilibrium point (K, 0). Setting S = α [1 − ( β / γ )], we obtain the solutions of
u 2 + Su + α = 0 as u* = [ − S ± S 2 − 4α ] / 2. Hence, v* = ( β / γK )[ K − u*]u * .
The non-trivial solutions exist if S 2 > 4α , and u* < K, that is if [1 − ( β / γ )]2 > ( 4 / α ), and
u* < K. If S < 0, that is β > γ , we obtain two positive equilibrium points. If S > 0, that is
β < γ , we have u* < 0, and there are no positive equilibrium points.
⎛ u⎞ v βu
10. We have F (u, v ) = ⎜1 − ⎟ − 2 , and G (u, v ) = 2 −γ .
⎝ K ⎠ (u / α ) + u + 1 (u / α ) + u + 1
(i) ( ∂F / ∂v ) < 0 , (v) F ( 0, 0) > 0 , (vi) F (0, A) = 0, A > 0 , (vii) F ( B, 0) = 0, B = K > 0,
are satisfied. Equality in condition (iii), ( ∂G / ∂v ) = 0 is satisfied.
17
 ⎛ ∂G ⎞ ⎛ ∂G ⎞
⎟ > 0, gives uβ [1 − (u / α )] > 0. Hence, we get the condition u < α .
2 2
(iv) u ⎜ ⎟ + v⎜
⎝ ∂u ⎠ ⎝ ∂v ⎠
⎛ ∂F ⎞ ⎛ ∂F ⎞
(ii) u ⎜ ⎟+v⎜ ⎟ < 0, gives (after simplification) the condition
⎝ ∂u ⎠ ⎝ ∂v ⎠
vK [( u 2 / α ) − 1] < u[( u 2 / α ) + u + 1]2 .
Since u 2 < α , the left hand side is negative and the inequality is satisfied as all other
quantities are positive.
(viii) G (C , 0) = 0, C > 0, gives the quadratic equation for C as γ C 2 + (γ − β )α C + αγ
= 0. The roots of the equation are 2C = [{( β / γ ) − 1} ± α {( β / γ ) − 1}2 − ( 4 / α ) ].
Both the roots are real and positive if β > γ , and {( β / γ ) − 1}2 > ( 4 / α ).
(ix) B > C gives K > C, that is K > (larger root of C).
Summarizing, we get the conditions as u 2 < α , β > γ , {( β / γ ) − 1}2 > ( 4 / α ),
K > 0.5 [{( β / γ ) − 1} + α {( β / γ ) − 1}2 − ( 4 / α ) ].
11. A stable equilibrium solution for the given model system exhibited for a typical set of
parameter values, K = 1, α = 3, β = 2.3 and γ = 0.3 is presented in Fig.2.3. We obtain
the non-zero equilibrium solution as (u*, v*) = (0.1511, 0.9836).
⎛ X⎞ BZ ⎛ w3 ⎞
12. We have F ( X , Z ) = A ⎜1 − ⎟ − , G ( X , Z ) = Z ⎜⎜ c − ⎟.
⎝ K ⎠ ( D + dX + Z ) ⎝ ( X + D3 ) ⎟⎠
Conditions (i), (ii), (v), (vii), (viii) are satisfied.
∂G w3 w − cD3
(iii) < 0, gives c − < 0, or X < 3 .
∂Z X + D3 c
Since X > 0, we obtain the condition c < ( w3 / D3 ).

Fig.2.3. Phase plot and time series for the model system (2.75)-(2.76) for K = 1,
α = 3, β = 2.3 and γ = 0.3.

⎛ ∂G ⎞ ⎛ ∂G ⎞ ⎡ w3 Z ⎤ ⎡ w3 ⎤
(iv) X ⎜ ⎟ + Z⎜ ⎟ > 0, gives X ⎢ 2⎥
+Z ⎢c − ⎥ >0.
⎝ ∂X ⎠ ⎝ ∂Z ⎠ ⎢⎣ ( X + D3 ) ⎥⎦ ⎣ X + D3 ⎦
⎡ w3 D3 ⎤ w3 D3
A sufficient condition is Z ⎢c − 2⎥
> 0, or c > .
⎢⎣ ( X + D3 ) ⎥⎦ ( X + D3 ) 2
18
 (vi) From F (0, A*) = 0, we obtain A* = AD /( B − A). The condition A* > 0, gives the
requirement B > A.
(ix) The condition B* > C *, gives K > C * . From (iii), we have C* < [( w3 − cD3 ) / c] .
We may choose K > [( w3 − cD3 ) / c] .
Summarizing the results, Kolmogorov theorem gives the following conditions.
w3 D3 w3
(a) Combining (iii), and (iv), we get 2
<c< .
( X + D3 ) ( X + D3 )
If X m is the maximum value of X, we can choose c < w3 /( X m + D3 ).
(b) B > A, (c) K > [( w3 − cD3 ) / c] .
⎛ Z ⎞ w3U wU
13. We have F ( Z ,U ) = A ⎜⎜1 − ⎟⎟ − , and G ( Z , U ) = c − 4 .
⎝ K1 ⎠ ( Z + D3 ) Z
Conditions (i), (ii), (iii), (v), (vi) are satisfied. Equality condition in (iv) is satisfied. The
requirement (vii) gives B* = K1 > 0. The requirement (ix) gives K1 > C * . But condition
(viii) is violated. We obtain G (C*, 0) = c. The condition G (C , 0) = 0, C > 0 is violated
since c ≠ 0. Hence, Kolmogorov theorem cannot be applied.
14. The oscillatory predator-prey dynamics exhibited by Holling-Tanner model (2.84), (2.85)
for the given set of parameter values is given in Fig.2.4.

Fig.2.4. Oscillatory predator-prey dynamics exhibited by Holling-Tanner model.

15. Note that (0, 0) is not an equilibrium point. (K, 0) is an equilibrium point. The second
equation gives U = cZ / w4 . Substituting in the first equation, we get
A w3cZ
(K − Z ) − = 0,
K α1 w4 + β1cZ + w4γ 1Z
or AKα1w4 + Z [ AK ( β1c + w4γ 1 ) − Aα1 w4 − w3 Kc] − AZ 2 ( β1c + w4γ 1 ) = 0,
α w + ( w3 Kc / A) Kα1 w4
or Z 2 + ( p − K ) Z − q = 0, where p = 1 4 ,q = .
( β1c + w4γ 1 ) ( β1c + w4γ 1 )
Irrespective of the sign of ( p − K ), the positive root is given by
2 Z * = ⎡− ( p − K ) + ( p − K ) 2 + 4q ⎤. We have U * = cZ * / w4 .
⎢⎣ ⎥⎦
An oscillatory predator-prey dynamics exhibited by the model system for the given set of
parameter values, A = 2, K = 100, w3 = 2.1, α1 = 0.45, β1 = 0.2, γ 1 = 0.6, c = 0.95 and
w4 = 1.65, is presented in Fig. 2.5.
19
 ⎛ Z⎞ w3U wU
16. We have F ( Z , U ) = A ⎜1 − ⎟ − , and G ( Z ,U ) = c − 4 .
⎝ K ⎠ (α1 + β1U + γ 1Z ) Z
Conditions (i), (ii), (iii), (v), (vii) are satisfied. Equality condition in (iv) is satisfied.
Aα1
(vi) F (0, A*) = 0, gives A* = > 0, if w3 > Aβ1 .
( w3 − Aβ1 )
(viii) G(C ,0) = c ≠ 0. The condition is not satisfied. (ix) B > C is also not satisfied .
Hence, Kolmogorov theorem cannot be applied.
17. The equilibrium point is X 1* = ( K1 − K 2 b1 ) /(1 − b1b2 ), X 2* = ( K 2 − K1b2 ) /(1 − b1b2 ).
Since X 1* > 0, and X 2* > 0, we obtain the conditions b1 < ( K1 / K 2 ) < (1 / b2 ), and
b1b2 < 1. The second condition is implied in the first condition. (Positivity holds also
when the inequalities are reversed). The elements of the Jacobian matrix are (dropping
the superfix *)
a11 = ( r1 / K1 )[ K1 − 2 X 1 − b1 X 2 ] = −( r1 X 1 / K1 ), a12 = −( r1b1 X 1 ) / K1 ,
a 21 = −( r2 b2 X 2 ) / K 2 , a 22 = ( r2 / K 2 )[ K 2 − b2 X 1 − 2 X 2 ] = −( r2 X 2 / K 2 ).
The characteristic equation is
λ2 + λ [( r1 X 1 / K1 ) + ( r2 X 2 ) / K 2 ] + [( r1 X 1 / K1 )( r2 X 2 ) / K 2 ](1 − b1b2 ) = 0.
Applying the Routh-Hurwitz criterion, we find that the positive equilibrium point is
asymptotically stable when b1b2 < 1. The required condition is b1 < ( K1 / K 2 ) < (1 / b2 ).

Fig.2.5. Time-series displaying oscillatory predator-prey dynamics

exhibited by the modified HT model (2.90)-(2.91).

18. The positive equilibrium point E * ( X 1* , X 2* ) is the solution of the equations

a1 − b1t1 − c1t 2 = 0, and a 2 − b2 t1 − c2 t2 = 0, where t1 = ln X 1 , t 2 = ln X 2 .
a c − a 2 c1 b a − b2 a1
We obtain t1 = 1 2 , t2 = 1 2 , X 1 = e t1 , X 2 = et 2 .
b1c2 − b2 c1 b1c2 − b2 c1
The elements of the Jacobian matrix are
a11 = a1 − b1 (1 + ln X 1 ) − c1 ln X 2 = −b1 , a12 = −c1 X 1 / X 2 ,
a 21 = −b2 X 2 / X 1 , a 22 = a 2 − b2 ln X 1 − c2 (1 + ln X 2 ) = −c2 .
The characteristic equation is λ2 + λ (b1 + c2 ) + (b1c2 − c1b2 ) = 0.

20
 Routh-Hurwitz criterion gives the necessary and sufficient conditions for the roots to be
negative or have negative real parts, as b1 + c2 > 0, and b1c2 − c1b2 > 0. The positive
equilibrium point is asymptotically stable when ( b1 / b2 ) > (c1 / c2 ).
For the given set of parameter values ( b1 / b2 ) = 3 / 4, and ( c1 / c2 ) = 2 / 3. The condition
is satisfied and the equilibrium point is asymptotically stable. The equilibrium point is
( X 1* , X 2* ) = ( e 7 , e −5 ).
19. For r = 1.5, α = 3, we get (1 / 3) < u* < 1. u* is a solution of
r (u * −1) 1
f (u*) = 1 + − e r ( u * −1) = 1.5 − − e1.5( u *−1) = 0.
αu * 2u *
Newton-Raphson’s method applied with the initial approximation taken as 0.5, gives the
sequence of iterates as 0.478602777, 0.480756737, 0.47971041, 0.480048641,
0.48004895. With u* = 0.48004895, we get v* = r (1 − u*) = 0.779926575. We obtain
p = 4.5u * −1.5 = 0.660220275, A + B + C = r (1 − u * p) = 1.0245929 > 0,
A − B + C = 2 − r + u * ( 2α − rp ) = 2.904886625 > 0,
A − C = 1 − u * (α − rp ) = 0.035260224 > 0.
By Miller’s theorem or Jury test, the equilibrium point is asymptotically stable.
20. The equilibrium points are obtained as (0, 0), and (4/9, 5/3). The elements of the Jacobian
matrix J are a11 = 2.5 − 3N − 0.5P, a12 = −0.5 N , a 21 = 1.8 P, a 22 = 0.2 + 1.8 N . At (0,
0), the eigen values of J are 2.5 and 0.2. The system is unstable. At (4/9, 5/3), the eigen
values of J are 1.0 and 1/3. The system is unstable.

Chapter 3
Exercise 3.1
1. MATLAB 7.0 is used to compute the phase plane diagram to generate the chaotic attractor
and time series. Chaotic attractor and the temporal evolution for (i) t vs x, (ii) t vs y, (iii) t
vs z are plotted in Figs. 3.1 (a), (b), (c) and (d).

11
0.8
10

9
0.6
z

8
0.4

7
0.8
0.6 1
0.2
0.8
0.4 0.6
0.2 0.4
0
0.2 0
y 0
x 4000 4500 5000 5500
t

(a) (b)

21
 0.5 11

0.4
10

0.3
9

z
y

0.2

8
0.1

0 7
4000 4500 5000 5500 4000 4500 5000 5500
t t

(c) (d)
Fig.3.1. (a) Chaotic attractor. (b) Temporal evolution (t vs x), (c) Temporal evolution
(t vs y), (d) Temporal evolution (t vs z), in Problem 1. (From Upadhyay, R. K., Rai, V.,
Complex dynamics and synchronization in two non-identical chaotic ecological systems. Chaos,
Solitons Fractals, 40, 2233–2241, Copyright 2009, Elsevier. Reprinted with permission.)

2. MATLAB 7.0 is used to generate the chaotic attractor in three different phase planes. The
projections of the chaotic attractor are drawn in Figs. 3.2. (a), (b), (c).

120 120

100 100

80 80

60 60
Z
Z

40 40

20 20

0 0
0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 45
X Y

(a) (b) (c)

Fig.3.2. Chaotic attractor: (a) (x-y) plane, (b) (x-z) plane, (c) (y-z) plane. Problem 2

1 b
0
0 1 2 3 4 5 6 7
0.8 -1

-2
0.6
log d
f

-3
0.4

-4
0.2
-5
0
0 1 2 3 4 5 6 7 -6
b

(a) (b)
Fig.3.3. Points in the 2D parameter spaces (a) (b, f ) , (b) (b, log d ). Problem 3.
(From Upadhyay, R. K., Kumari, N., Rai, V., Exploring dynamical complexity in diffusion
driven predator–prey systems: Effect of toxin production by phytoplankton and spatial
heterogeneities. Chaos, Solitons Fractals, 42(1), 584–594. Copyright 2009, Elsevier. Reprinted
with permission).

3. MATLAB 7.0 is used to compute the discrete points at which chaos was observed and
Microsoft office Excel 2007 to draw the 2D Scan diagram. All the points in the two
dimensional parameter spaces where the model system exhibits chaotic dynamics are
shown in Figs. 3.3(a), (b).
22
 dx wxy dy w1 xy
4. The subsystem I is = a1 x − b1 x 2 − , = −a2 y + .
dt (α + β y + γ x) dt (α + β y + γ x)
The local stability conditions are γw − βw1 < 0, a2 ( β w1 − γ w) − ww1 < 0, ( w1 − γa 2 ) > 0,
γ a1 a2
and < b1 x < a1 .
( w1 + γ a2 )
a a2α
The Kolmogorov conditions give w > a1 β , w1 > a2 γ and 1 > .
b1 ( w1 − a2 γ )
A set of parametric values satisfying the Kolmogorov conditions and violating the local
stability condition are a1 = 2, b1 = 0.06, w = 0.85, α = 0.45, β = 0.2, γ = 0.6, a2 = 1.05,
and w1 = 1.65. Now, assume that there exists a prey x, in addition to y whose growth rate
is governed by
dx ⎛ x ⎞ 
Bxz
= Ax ⎜1 − ⎟ − ,
dt ⎝ K ⎠ (α + β z + γ x )
where A is the rate of self-reproduction for this prey and K is the carrying capacity of its
environment. In the Leslie-Gower scheme, the growth rate equations for the two
populations will give the subsystem II
dx ⎛ x ⎞ 
Bxz dz w z2
= Ax ⎜1 − ⎟ − , = cz 2 − 3 .
dt ⎝ K ⎠ (α + β z + γ x ) dt x + D3
For this subsystem to be a K-system, the following conditions are to be satisfied
w3 D3 w3
(i) <c< . If xm is the maximum value of x, we can choose
( x + D3 ) 2
( x + D3 )
c < w3 /( xm + D3 ).
(ii) w3 > cD3 , (iii) K > [( w3 − cD3 ) / c] .
Combining the above two conditions, we obtain
D3 w3
< c( x + D3 ) < w3 < c( K + D3 ); cD3 < w3 (3.1)
( x + D3 )
The subsystem II exhibits a stable equilibrium when the following condition is satisfied:
w3 A(1 − x * / K )(α + γ x * )
KBγ z * < A(α + β z* + γ x * ) 2 , x * = − D3 , z * = (3.2)
c {B − Aβ (1 − x* / K )}
When the values of the parameters of this subsystem II are chosen in such a way
that the constraints (3.1) are satisfied and the inequality (3.2) is violated, the subsystem II
admits a limit cycle solution. For example, for the set of parametric values,
A = 2.0, K = 130, α = 0.45, β = 0.2, γ = 0.6, B = 0.74, c = 0.0295, w3 = 2.1, D3 = 20 ,
the subsystem II has limit cycle solutions. The linking scheme would depend on how the
individual populations of the two subsystems are related with each other. The link scheme
can be mathematically represented by adding the term {−[( w2 yz ) /(α + β y + γx )]} to the
second equation of subsystem I. The selection of values for w2 is based on the fact that it
plays a role similar to that of w. w2 can be varied from 0.1 to 1. The parametric values for
the original system are selected by omitting those appearing in the growth equation of
pseudo prey ( x ). A set of parameter values for which the system admits limit cycle
23
 solution is a1 = 2, b1 = 0.06, w =0.85, a2 = 1.05, w1 = 1.65, w2 = 0.3, α = 0.45, β = 0.2,
γ = 0.6, c = 0.0295, w3 =2.1, D3 = 20. There exist other sets of parameter values which
satisfy the above criteria.
5. BAS routine from Dynamics: Numerical Explorations software is used to compute the
basin boundary structure, which bring changes in the dynamical behavior when the system
parameters are varied . Basin boundary structure is plotted in Fig.3.4.

Fig.3.4. Basin boundary structure for the model given in Problem 5.

(From Upadhyay, R. K., Rao, V. S. H., Short term recurrent chaos and role of toxin
producing phytoplankton on chaotic dynamics in aquatic systems. Chaos, Solitons Fractals,
39, 1550–1564. Copyright 2009, Elsevier. Reprinted with permission).

6. MATLAB 7.0 is used to generate the bifurcation diagrams. The bifurcation diagrams are
given in Fig. 3.5. From the figures, transition from chaos to order through a sequence of
period halving bifurcation and hence a period-doubling cascade can be observed.

7.MATLAB 7.0 is used to generate the bifurcation diagrams. The successive maxima of y as
a function of w5 for the given parameter values with w11 = 0.03 and 0.15 < w5 < 0.5, are
plotted in Fig.3.6 (i); and with w11 = 0.06 and 0.25 < w5 < 0.5, are plotted in Fig.3.6 (ii).
8. We have f (0, 0) = 0, ( ∂f / ∂x )(0, 0) = 0, ( ∂f / ∂μ )(0, 0) = 1 / 4, ( ∂ 2 f / ∂x 2 ) ( 0, 0) = −2.
For, μ < 0, the model has no fixed points and the vector field is decreasing in x . For,
μ > 0, the model has two fixed points x = ± μ / 2. The fixed point x = μ / 2 is stable,
and the second fixed point x = − μ / 2 is unstable.

9. The nullclines intersect when x[ μax 2 − x + 9aμ ] = 0. The fixed points are (0, 0), and
x* = x1,2 = [1 ± 1 − 36a 2 μ 2 ] /(2aμ ). These two solutions coincide when 6aμ = 1 . Hence,
the critical value is μc = 1/ 6a. At the bifurcation, the value of the fixed point is x* = 3.
The Jacobian matrix is
⎡ −μ 1 ⎤
⎥. Trace(J) = − ( a + μ ). J = aμ − [18 x /(9 + x ) ].
2 2
J =⎢
⎣18 x /[( 9 + x 2 2
) ] − a ⎦
All the fixed points are either sinks or saddle points. At (0, 0), J = aμ > 0. Hence, (0, 0)
is always a stable fixed point. At the other two fixed points, we have

24
 Fig.3.5. Bifurcation diagrams taking θ as a control parameter for the model given in
Problem 6. (From Upadhyay, R. K., Niji, R. K., Nitu Kumari. Dynamical complicity in some
ecological models: Effect of touin production by phytoplankton. Nonlinear Andy.: Model.
control 12(1), 123–138. Copyright 2007, Lithuanian Association of Nonlinear Analysts
(LANA). Reprinted with permission).

(i) (ii)

Fig.3.6. Bifurcation diagram of the model system. Problem 7.

(From Upadhyay, R. K., Raw, S. N., Rai, V. 2010. Dynamical complexities in a tri-trophic hybrid
food chain model with Holling type II and Crawley–Martin functional responses. Nonlinear
Anal.: Model. Control 15(3), 361–375. Copyright © 2010, Lithuanian Association of Nonlinear
Analysts, (LANA). Reprinted with permission).

18 x 18 x ⎡ x2 − 9 ⎤
J = aμ − = aμ − = aμ ⎢ 2 ⎥.
(9 + x 2 ) 2 (9 + x 2 )( x / μa ) ⎣ x + 9 ⎦
For 0 < x* < 3, J < 0. The fixed point is a saddle point. For x* > 3, J > 0. The fixed
point is a stable node.
10. We have f (0, 0) = 0, ( ∂f / ∂x )(0, 0) = 0, ( ∂f / ∂μ )(0, 0) = 0, ( ∂ 2 f / ∂x 2 ) ( 0, 0) = 8,
( ∂ 2 f / ∂x∂μ )( 0, 0) = 1. Conditions (3.25) are satisfied. The system under-goes transcritical
25
 bifurcation at (0, 0). Fixed points are x = 0 and x = 2 ± 4 + μ . Now, ( ∂f / ∂x )( 0, μ ) = μ ,
which changes sign as μ goes through zero. The system changes its stability at (0, 0). For
x2 = 2 − 4 + μ , we have x2 > 0 for μ < 0, and x2 < 0 for μ > 0, and ( ∂f / ∂x )( x2 , μ )
= −4 x2 − 2 μ changes its sign from positive to negative as μ goes through zero.

11. (i) We have f (0, 0) = 0, ( ∂f / ∂x )(0, 0) = 0, ( ∂f / ∂μ )(0, 0) = 0, ( ∂ 2 f / ∂x 2 )(0, 0) = 0,

( ∂ 2 f / ∂x∂μ )(0, 0) = 1, ( ∂ 3 f / ∂x 3 )( 0, 0) = −6. The fixed points are x = 0 and x = ± μ .
For μ < 0 , the fixed point x = 0 is stable. For μ > 0 , all the three fixed points exist.
x = 0 is unstable and the other two fixed points are stable giving rise to supercritical
pitchfork bifurcation.
(ii) We have f (0, 0) = 0, ( ∂f / ∂x )( 0, 0) = 0, ( ∂f / ∂μ )(0, 0) = 0, ( ∂ 2 f / ∂x 2 )(0, 0) = 0,
( ∂ 2 f / ∂x∂μ )(0, 0) = 1, ( ∂ 3 f / ∂x 3 )( 0, 0) = 6. The fixed points are x = 0 and x = ± − μ .
For μ < 0 , the fixed point x = 0 is stable and x = ± − μ are unstable. The stationary
solution (node) becomes an unstable saddle (a saddle-node, a saddle-focus), and together
with it, the other two unstable stationary solutions disappear. The bifurcation is a crisis
giving rise to subcritical pitchfork bifurcation.
12. We have f (0, 0) = 0, ( ∂f / ∂x )(0, 0) = 1, ( ∂f / ∂μ )( 0, 0) = 1 / 4, ( ∂ 2 f / ∂x 2 ) ( 0, 0) = −2.
The map satisfies the conditions (3.24) and Remark 3.3. The sign of [ −(∂ 2 f / ∂x 2 )( 0, 0) /
( ∂f / ∂μ )( 0, 0)] = 8, tells us on which side of μ = 0 , the curve of fixed points is located.
The map has no fixed points for μ < 0. For μ > 0, the map has two fixed points
x = ± μ / 2 . The fixed point x = μ / 2 is stable and x = − μ / 2 is unstable. There are
no fixed points on one side of μ = 0 and two fixed points on the other side. Saddle node
bifurcation occurs at ( x, μ ) = (0, 0).
13. We have f (0, 0) = 0, ( ∂f / ∂x )(0, 0) = 1, ( ∂f / ∂μ )(0, 0) = 0, ( ∂ 2 f / ∂x∂μ ) ( 0, 0) = 1 / 2,
( ∂ 2 f / ∂x 2 ) ( 0, 0) = −2. Fixed points are x = 0 and x = μ / 2 , For μ < 0, x = 0 is stable
and x = μ / 2 is unstable. For μ > 0, x = 0 is unstable and x = μ / 2 is stable. Thus, an
exchange of stability occurs at μ = 0 with respect to both the fixed points. Transcritical
bifurcation occurs at ( x, μ ) = (0, 0).
14. We have f (0, 0) = 0, ( ∂f / ∂x )(0, 0) = 1, ( ∂f / ∂μ )(0, 0) = 0, ( ∂ 2 f / ∂x∂μ ) ( 0, 0) = 1 / 4,
( ∂ 2 f / ∂x 2 )(0, 0) = 0, ( ∂ 3 f / ∂x 3 )( 0, 0) = −6. The fixed points of the map are x = 0 and
x = ± μ / 2. For μ < 0, the fixed point x = 0 is stable. For μ > 0, the fixed point x = 0
is unstable and other two fixed points are stable. Pitchfork bifurcation occurs at ( x, μ ) =
(0, 0).
15. The fixed point x0 = [ −1 + 1 + 4 μ ] / 2 > 0 is called a period-1 cycle. It is stable for
μ ∈ ( 0, 3 / 4] and its multiplier is − 2 x0 . At μ = μ1 = 3 / 4, the multiplier is −1 and the
first period doubling bifurcation takes place giving rise to a stable period-2 cycle. The 2-
cycle consists of two points x1,2 = [1 ± 4 μ − 3 ] / 2. The 2-cycle is stable in the interval
μ ∈ [3 / 4, 5 / 4] and has the multiplier f x ( x1 ) f x ( x2 ) = 4(1 − μ ). At μ = μ2 = 5 / 4, the
26
 second period doubling bifurcation takes place giving rise to a stable period- 2 2 cycle.
The accumulation point of the bifurcation points is μ c = 1.40115...
⎡ μ − 2⎤
16. We have at (0, 0), J = ⎢ ⎥. det( J ) = μ 2 + 1 > 0. Trace( J ) = 2μ . Eigenvalues are
⎣1 / 2 μ ⎦
λ = μ ± i. As μ increases through zero, origin changes from a stable spiral to an
unstable spiral. Hopf bifurcation occurs at μ = 0. Since an unstable limit cycle surrounds
the stable fixed point, it is subcritical.

17. The variational matrix for E * = ( X ∗ , Y ∗ , Z ∗ ) is given by

⎡ a11 a12 0 ⎤
⎢
V4 = ⎢a 21 a 22 a 23 ⎥⎥ I have not checked as it is a paper. Is it OK.
⎢⎣ 0 a32 0 ⎥⎦
⎛ γ wY * ⎞⎟ wX * (α + γ X *) w1 ( α + β Y * )Y ∗
where a11 = X * ⎜ − b1 + , a = − , a = ,
⎜ 2 ⎟ 12 2 21 2
⎝ S1 ⎠ S1 S1
⎛w β X ∗
w γ Z ⎞ ∗ ∗
w Y (α 2 + γ 2Y )∗
a 22 = − Y * ⎜ 1 2 − 2 22 ⎟, a 23 = − 2 ,
⎜ S ⎟ 2
⎝ 1 S 2 ⎠ S 2

a32 = c 2 Z *2 / w3 , S1 = (α + βY * +γ X *), S 2 = (α 2 + β 2 Z * +γ 2Y *).

γ wY *
Note that a12 < 0, a 21 > 0, a 23 < 0, a32 > 0. Also, a11 < 0, if < b1 ,
S12
w2 γ 2 Z * w1 β X *
and a 22 < 0, if < .
S 22 S12
The characteristic equation is given by λ 3 + p2 λ 2 + p1 λ + p0 = 0 , where
p2 = −(a11 + a22 ) , p1 = a11a22 − a23 a32 − a12 a21 , p0 = a11 a23 a32 .
Hopf bifurcation occurs at μ = μ 0 , if
p0 ( μ0 ) > 0, D1 ( μ0 ) = p1 ( μ0 ) > 0, D2 ( μ0 ) = p1 ( μ0 ) p2 ( μ0 ) − p0 ( μ0 ) = 0,
and [ dD2 ( μ0 ) / dμ ] ≠ 0.
Let, c, the growth rate of the generalist predator be a bifurcation parameter. Now, p2
is independent of c. From the above analysis, if a 22 ≤ 0 , along with the other conditions,
then p1 p2 − p0 > 0 . Therefore, we assume that a 22 > 0. That is,
w2γ 2 Z * w1 β X *
> . (3.3)
(α 2 + β 2 Z + γ 2Y ) 2
(α + β Y + γ X ) 2
Now, p2 = −(a11 + a22 ) > 0 , if
Y * ⎡ ( wγ − w1 β ) X ∗ w γ2Z∗ ⎤
⎢ + 2 ⎥
< b1 . (3.4)
X * ⎣ (α + β Y + γ X ) 2
(α 2 + β 2 Z + γ 2Y ) ⎦
27
 γ wY *
p0 = a11 a23 a32 > 0 , if a11 < 0. Now, a11 < 0, if < b1 is satisfied.
(α + β Y + γ X ) 2
Solving p1 p2 − p0 = ( a11 + a22 ) (a12 a21 − a11 a22 ) + a22 a23 a32 = 0 for c, we obtain
a 22 (− a 23 )(c 2 Z * 2 / w3 ) = ( a11 + a 22 )( a12 a 21 − a11a 22 ).
The left hand side is positive. We require that the right hand side be positive so that we
have a positive solution for c. Since (a11 + a 22 ) < 0, a sufficient condition is
( a12 a 21 − a11a 22 ) < 0. We require
⎛ γ wY * ⎞⎛ w2 γ 2 Z ∗ w1 β X ∗ ⎞
⎜ 1
b − 2 ⎟ ⎜
− 2 ⎟
⎝ (α + β Y + γ X ) ⎠ ⎝ (α 2 + β 2 Z + γ 2Y ) 2
(α + β Y + γ X ) ⎠
ww1 (α + γ X *)( α + β Y )*
< . (3.5)
(α + β Y + γ X ) 4
Therefore, c exists and (dD2 / dc) ≠ 0. The conditions to be satisfied are (3.3), (3.4), (3.5).
18. The Lyapunov exponent bifurcation diagrams are drawn using MATLAB 7.0. The
diagrams are plotted in Fig.3.7 for (i) ω12 = 0.25, 0.01 ≤ ω8 ≤ 0.6; and (ii) ω8 = 0.21,
and 0.15 ≤ ω12 ≤ 0.3.

Fig.3.7. Lyapunov exponent bifurcation diagram as a function of (i) ω8 , and

(ii) ω12 , Problem 18.
19. We obtain f ′( x ) = 2. We may expect that all orbits will be unstable and chaos sets in.
We obtain
1 n 1 n
λ = lim log ∏ f ′( xi ) = lim log ∏ ( 2) = log 2.
n→∞ n i =1 n→∞ n i =1
20. We solve the system of equations using MATHEMATICA. The solutions are given in
Table 3.1. The corresponding variational equations are
dp dq
= (4.5 − 0.25 F ) p − 0.25 Rq, = 0.5 Fp + q (0.5 R − 4).
dt dt
Computation of Lyapunov Exponents of order 1 The variational equations are also solved
using MATHEMATICA. Assume the initial vectors as ω1 (0) = (1/ 2, 1/ 2)T , and
ω2 (0) = (1, 0)T , for computing λ1, λ2 respectively. At every time step, denote the
normalized ω by ω̂. The numerical solutions obtained for computing λ1 are given in Table
3.2 and for computing λ2 are given in Table 3.3.

28
 Table 3.1. Solutions of Lotka-Volterra model. Problem 20.

t R(t) F(t) t R(t) F(t)

0.1 12.0008 11.6259 0.6 7.00788 30.1885
0.2 13.573 14.8196 0.7 5.32247 27.4408
0.3 13.8507 19.854 0.8 4.41693 23.3969
0.4 12.2791 25.7597 0.9 4.06403 19.3467
0.5 9.54427 29.8573 1.0 4.11015 15.8837
Table 3.2. Solutions for computing λ1 , Problem 20.

t 0.1 0.2 0.3 0.4 0.5

ω1 (t ) ⎛ 0.609011 ⎞ ⎛ 0.377234 ⎞ ⎛ 0.001498 ⎞ ⎛ −0.505948 ⎞ ⎛ −1.10702 ⎞
⎜1.1332 ⎟ ⎜1.51901 ⎟ ⎜1.78594 ⎟ ⎜1.84843 ⎟ ⎜1.62439 ⎟
⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠
t 0.6 0.7 0.8 0.9 1.0
ω1 (t ) ⎛ −1.73357 ⎞ ⎛ −2.28991 ⎞ ⎛ −2.6586 ⎞ ⎛ −2.71242 ⎞ ⎛ −2.3322 ⎞
⎜1.05079 ⎟ ⎜ 0.101106 ⎟ ⎜ −1.19816 ⎟ ⎜ −2.75219 ⎟ ⎜ −4.39071 ⎟
⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠

Table 3.3. Solutions for computing λ2 , Problem 20.

t 0.1 0.2 0.3 0.4 0.5

ω1 (t ) ⎛ 1.14687 ⎞ ⎛ 1.155 ⎞ ⎛ 0.973051 ⎞ ⎛ 0.564971 ⎞ ⎛ − 0.0803505 ⎞
⎜ 0.57001 ⎟ ⎜1.24213 ⎟ ⎜1.94191 ⎟ ⎜ 2.5627 ⎟ ⎜ 2.97288 ⎟
⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠
t 0.6 0.7 0.8 0.9 1.0
ω1 (t ) ⎛ −0.93794 ⎞ ⎛ −1.93921 ⎞ ⎛ −2.96714 ⎞ ⎛ −3.85943 ⎞ ⎛ −4.41958 ⎞
⎜ 3.03038 ⎟ ⎜ 2.60177 ⎟ ⎜ 1.58855 ⎟ ⎜ −0.0446107 ⎟ ⎜ −2.24203 ⎟
⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠

Using (3.31), at t = 1 , we obtain λ1 = 8.16805, and λ2 = 10.1024. The maximal Lyaponov

exponent is 10.1024 > 0. The plot of λ1 , λ2 is given in Fig.3.8.

Fig. 3.8. Lyaponov exponents (LCE of order 1).

29
 Chapter 4
Exercise 4.1

1. The non-dimensionalized form is

∂p ⎡ az ⎤ ∂2 p ∂z ⎡ np f z ⎤ ∂2z
= p ⎢rx (1 − p ) − + d , = z⎢ −m− ⎥ + d ,
∂t ⎣ 1 + bp ⎥⎦ ∂x 2 ∂t ⎣1 + bp 1 + g 2z2 ⎦ ∂x 2
where
a = K /( R H P ), b = K / H P , d = D /( L2 R ), m = δ / R , n = eυa = eυK /( R H P ), f = FK
/(υR H Z2 ), g = K /(υH Z ), rx = ( R x / R ) = s + lx. Let ( p * , z * ) be the positive equilibrium
solution of the system in the absence of space and diffusion. Writing
p = p * + U , z = z * + V , we obtain the linearized form of the model system about the
positive equilibrium point E * ( p* , z * ) as
∂U ∂ 2U ∂V ∂ 2V
= a11U + a12V + d 2 , = a21U + a 22V + d 2 ,
∂t ∂x ∂t ∂x
* * *
az abp z ap *
where a11 = rx (1 − 2 p * ) − = − p *
rx + , a 12 = − ,
(1 + bp* ) 2 (1 + bp* ) 2 1 + bp*
nz* np * 2 fz * fz* ( g 2 z *2 − 1)
a 21 = , a 22 = −m− = .
(1 + bp * ) 2 (1 + bp* ) (1 + g 2 z *2 ) 2 (1 + g 2 z *2 ) 2
Write the solution of the equations in the form U = se λt + ikx , V = we λt + ikx , where λ
and k are the frequency and wave number respectively. Substitute the expressions for U, V
in the equations The homogeneous equations in s and w have solution if determinant of the
coefficient matrix is zero. We get
( λ − a11 + dk 2 )( λ − a 22 + dk 2 ) − a12 a21 = 0, or λ2 + p1λ + q1 = 0,
where p1 = 2dk 2 − ( a11 + a 22 ), q1 = d 2 k 4 − dk 2 ( a11 + a 22 ) + ( a11a 22 − a12 a 21 ).
By Routh-Hurwitz criterion, the roots of the characteristic equation are negative or
have negative real parts if p1 > 0, and q1 > 0. A sufficient condition for p1 > 0 is
a11 + a 22 < 0. This condition is satisfied when abz* < rx (1 + bp* ) 2 , and g 2 z*2 < 1. For
these values, a11 < 0, a 22 < 0. Also, a12 < 0 and a 21 > 0. Hence, q1 > 0. Therefore, the
positive equilibrium E * is locally asymptotically stable in the presence of diffusion if the
above two sufficient conditions are satisfied. The positive equilibrium point E * is also
locally asymptotically stable in the absence of diffusion. However, irrespective of the sign
of p1 , the system is unstable if q1 < 0, as the characteristic equation has a positive root.
This implies that there exist parameter values and a range for d, for which the system may
be unstable.
2. Numerical simulations are done using MATLAB 7.0. For the numerical integration,
Runge-Kutta fourth order method was used. Using the given values of the parameters, we
have rx = 2 − 1.4 x. Now, ( p * , z * ) is the solution of the equations s(1 − p ) − [az /(1 + bp)]
= 0, [np /(1 + bp )] − m − [ fz /(1 + g 2 z 2 ] = 0. With the given set of values, we obtain using
Newton’s iteration, p * = 0.1714390348, z * = 0.473 4767465. Computations are done
30
 with the two given sets of initial conditions to study the sensitivity of the solutions to
changes in the initial conditions. The results show that the prey density of model system is
sensitive to initial condition at x = 0.85 , ( rx = 0.81) (see Fig. 4.1).

The irregular spatial and temporal behaviors of the prey and predator densities of the
model system are plotted in Figs. 4.2. Space-time plots of the prey and predator densities
display irregular and complex dynamics in the model system. These spatiotemporal
patterns reflect the effect of environmental heterogeneity at both spatial and temporal
scales. Highly disordered oscillations are observed in both space and time. We observe that
the fish predation stabilizes the dynamics of model system.

Fig. 4.1. Sensitivity of prey density to initial conditions (time series).

Fig.4.2. Complex spatiotemporal patterns of prey density and predator density. (From
Upadhyay, R. K., Kumari, N., Rai, V. Wave of chaos in a diffusive systems: Generarting
realistic patterns of patchiness in plankton-fish dynamics. Chaos, Solitons Fractals, 40(1), 262–
276. Copyright 2009, Elsevier. Reprinted with permission).

3. Perturb the solution of (4.38), (4.39) about the steady state ( P* , H * ) to ( P * + U ,

H * + V ), where U, V are small. The linearized system is given by
∂U ∂ 2U ∂H ∂ 2V
= a11U + a12V + d1 2 , = a21U + a22V + d 2 2 . (4.1)
∂t ∂x ∂t ∂x

31
 2 B2 P*3 H * B2 P *2
where a11 = − r + , a12 = − ,
( P *2 + D 2 ) 2
P *2 + D 2
H *2C2 2 FH *3
a21 = , a22 = −C1 + ,
P *2 ( H *2 + D12 ) 2

(the equations governing the equilibrium point are used). Assume the solution of the above
equations in the form U ( x, t ) = seλt cos(kx ), V ( x, t ) = we λt sin(kx), where k takes the
discrete values, k = ( nπ L), n is an integer and λ is the eigen value determining the
temporal growth (frequency). Here, k denotes the wave number. Substitute the expressions
for U and V in (4.1). The homogeneous equations in s and w have a solution if the
determinant of the coefficient matrix is zero. We obtain the characteristic equation as
λ2 + pλ + q = 0, where p = ( d1 + d 2 )k 2 − (a11 + a 22 ), q = d1d 2 k 4 − ( a11d 2 + a 22 d1 )
k 2 + a11a 22 − a12 a 21 . Note that a12 < 0, a 21 > 0. Now, a sufficient condition for p > 0 is
a11 + a22 < 0, that is,
*⎡ B2 P*3 FH *2⎤
r + C1 > 2 H ⎢ *2 2 2
+ ⎥. (4.2)
⎣(P + D ) ( H *2 + D12 ) 2 ⎦
If q > 0 is also satisfied, then both the eigenvalues are negative or have negative real parts.
The steady state ( P* , H * ) is linearly stable both in the presence as well as in the absence of
diffusion. For the parameter values, r = 1, B1 = 0.2, B2 = 0.91, D 2 = 0.3, C1 = 0.22, C 2 =
0.2, D1 = 0.1, F = 0.02, the equilibrium point is obtained as P* = 1.184 922937, and H * =
1.205819205. Without diffusion, we obtain the characteristic equation as λ2 − 0.07009λ
+ 0.10709 = 0. The eigenvalues are a complex pair with positive real parts. In the absence
of diffusion, the system is unstable. In order to have Turing instability, we require that one
or both the eigen values have positive real parts for some k ≠ 0 . Both the roots are positive
or have positive real parts if p < 0, and q > 0. One of the roots is positive if p > 0, and
q < 0; or when p < 0, and q < 0. Now, p > 0, for all d1 , d 2 , when the sufficient condition
(4.2) is satisfied. Diffusive instability can arise if q < 0, that is if q( k 2 ) = Dk 4 − k 2C + B
< 0, where D = d1d 2 , B = a11a22 − a12 a21, and C = (d 2a11 + d1a22 ). The roots of this
equation in k 2 are real and positive when (i) B > 0, (ii) C > 0, and (iii) C 2 − 4 BD > 0.
Then, q < 0 when k12 < k 2 < k22 , where k12 , k 22 = [C  C 2 − 4 BD ] /(2 D ). Therefore,
diffusive instability occurs when these conditions and (4.2) are satisfied. Now, q(k 2 ) is a
quadratic in k 2 and the graph of y = q( k 2 ), is a parabola opening upwards. The minimum
occurs at the vertex of the parabola, that is for k 2 = km2 where k 2 = k m2 = C /(2 D).
Consider again the above parameter values. The condition (4.2) is not satisfied. We obtain,
B = 0.1070942, C > 0 when ( d 2 / d1 ) > 0.7276648; C 2 − 4 BD > 0 when ( d 2 / d1 ) > 7.8551.
Let, d1 = 1, d 2 = 8. Then, p > 0 when k 2 > 0.00779. Now, q > 0 when k 2 > 0.134231. For
all wave numbers k 2 > 0.134231, the system is stable under the action of diffusion. Note

32
 that when there is no diffusion, the equilibrium point is unstable. When 0.09973 < k 2 <
0.13423, p > 0, and q < 0. For all these wave numbers, diffusive instability sets in.
4. For the one-dimensional case, the temporal dynamics is studied by observing the effect of
time on space vs. density plot of prey populations. The spatiotemporal dynamics of the
system depends to a large extent on the choice of initial conditions. In a real aquatic
ecosystem, the initial spatial distribution of the species can be caused by spatially
homogenous initial conditions. However, in this case, the distribution of species would stay
homogenous for any time, and no spatial pattern can emerge. To get nontrivial
spatiotemporal dynamics, the homogenous distribution is to be perturbed. We consider the
constant-gradient distribution [54]: P( x, 0) = P* , H ( x, 0) = H * + εx + δ , where ε = −1.5 ×
10−6 , δ = 10−6 are parameters and ( P * , H * ) ≈ (1.8789, 1.3984) is the non-trivial
equilibrium point for the parameter values r = 1, B1 = 0.2, B2 = 0.91, D 2 = 0.3, C1 =
0.22, C2 = 0.2, D1 = 0.1, F = 0.1. Population distributions over space at t = 2000 are
given in Fig.4.3a, and with ε = −1.5 × 10 −3 , δ = 10−6 are given in Fig.4.3b. In Fig.4.3a, the
spatial distributions gradually vary in time, and the local temporal behavior of the
dynamical variables P and H is strictly periodic following limit cycle of the non-spatial
system, which is perhaps intuitively expected from the one dimensional spatial model
system. When ε in increased in magnitude from − 1.5 × 10 −6 to − 1.5 × 10−3 , the dynamics
of the system undergoes major qualitative changes (see Fig.4.4b). The size of the region
occupied by this pattern steadily grows with time, so that finally irregular spatiotemporal
oscillations invade the whole domain. Population distributions for ε = −1.5 × 10 −2 ,
δ = 10 −3 and over space at (a) t = 1000 and (b) t = 2000 are plotted in Figs.4.4a,b.
(a) (b)

Fig.4.3. Population distributions over space at t = 2000.

(From Upadhyay, R. K., Wang, W., Thakur, N. K. Spatiotemporal dynamics in a plankton
system. Math. Model. Nat. Phenom. 5(5), 101–121. Copyright 2010, Cambridge University
press(EDP Science). Reprinted with permission).

For the two-dimensional case, Turing patterns are obtained by performing numerical
simulations with system size 200 × 200 for different values of F and time t. The values of the
parameters are same as in the one dimensional case. The values of the other parameters are
D1 = 0.01, d1 = 0.05, d 2 = 1. The uniform mesh size along both axis is taken as h = 1/3, and
the time step is taken as Δt = 1 / 40. Initially, the entire system is placed in the stationary state
( P* , H * ) and the propagation velocity is of the order of 5 × 10 −4 space units per time unit.

33
 (a) (b)

Figs.4.4. Population distributions over space at (a) t = 1000 and (b) t = 2000.
(From Upadhyay, R. K., Wang, W., Thakur, N. K. Spatiotemporal dynamics in a plankton
system. Math. Model. Nat. Phenom. 5(5), 101–121. Copyright 2010, Cambridge University
press(EDP Science). Reprinted with permission).
Then, the system is integrated for 105 or 2 × 105 time steps and some images are saved.
After the initial period during which the perturbation spreads, the system goes either into a
time dependent state or to an essentially steady state. As an illustration, the typical Turing
patterns of phytoplankton P are plotted in Figs.4.5a-e. We observe that the patterns change as
F is increased. For F = 0.0001, 0.01, 0.035, 0.065, 0.0768, we observe spots pattern, spot-
stripe mixtures pattern, stripes pattern, hole-stripe mixtures, and holes pattern respectively.
For F = 0.035, the system is integrated for 105 time steps and for other cases for 2 × 105 time
steps. For F > 0.077, wave pattern emerges.

Fig.4.5a-e. Typical Turing patterns of P in the two dimensional spatial model system.
(From Upadhyay, R. K., Wang, W., Thakur, N. K. Spatiotemporal dynamics in a plankton
system. Math. Model. Nat. Phenom. 5(5), 101–121. Copyright 2010, Cambridge University
press(EDP Science). Reprinted with permission).

34
5. From the first equation, we get b1 x = a1 − [ wy /( x + D )]. Hence, x ≤ a1 / b1. Equilibrium
points of the system are E0 (0, 0, 0), E1 ( a1 / b1 , 0, 0), E2 ( xˆ , yˆ , 0), where x̂ is the positive
solution of ( xˆ + D1 )[ a 2 + θ f ( xˆ )] − w1 xˆ = 0, yˆ = ( a1 − b1 xˆ )( xˆ + D ) / w, xˆ < a1 / b1 . The
positive equilibrium point is E3 ( x* , y * , z * ) , where y * = ( w3 − cD3 ) / c, w3 > cD3 , x* is
the positive solution of the equation b1 x 2 + ( Db1 − a1 ) x + ( wy * − a1 D ) = 0, z* =
( y * + D2 )[ w1 x * − ( x* + D1 )( a 2 + θ f ( x* ))] /[ w2 ( x * + D1 )]. Since, z * > 0, we get the
condition w1 x * > ( x * + D1 )( a 2 + θ f ( x * )). For E0 (0, 0, 0), the eigenvalues of the Jacobian
matrix are 0, a1 , − a 2 , for f (0) = 0. E0 (0, 0, 0) is a non-hyperbolic fixed point. Further-
more, as one eigenvalue is positive and another one is negative, E0 is always non-stable.
For E1 ( a1 / b1 , 0, 0), the eigenvalues of the Jacobian matrix are 0, − a1 , − a 2 − θ f ( a1 / b1 )
+ [ w1a1 /(a1 + b1 D1 )]. If [ w1a1 /(a1 + b1 D1 )] < a 2 + θ f (a1 / b1 ), the third eigenvalue is
negative. E1 is also a non-hyperbolic fixed point. Since two of the eigen values are
negative, E1 has a stable manifold in at least two dimensions. To know the actual
dimensions of the stable and unstable manifolds of E 1 we need to compute the center
manifold corresponding to the eigenvalue zero [1]. For E2 ( xˆ , yˆ , 0), one eigenvalue of the
Jacobian matrix is zero and the other eigenvalues are root of a quadratic equation. The
Jacobian of the linearized systems about E0 , E1 , and E 2 has zero as one of the
eigenvalues. These are non-hyperbolic points and the dynamical behavior near them can be
stable, periodic, or even chaotic. At E3 ( x* , y * , z * ) , the characteristic equation of J is
λ3 + s1λ2 + s2 λ + s3 = 0, where s1 = −( a11 + a 22 ), s2 = ( a11a 22 − a12 a 21 − a 23a 32 ),
s3 = a11a 23a 32 , a11 = x * [ −b1 + {wy * /( x * + D ) 2 }], a12 = − wx * /( x * + D ), a 21 = [{w1 D1 y *
/( x* + D1 )2 } − θy * f ′( x* )], a 22 = w2 y * z * /( y * + D2 ) 2 , a23 = − w2 y * /( y * + D2 ), a 32 =
w3 z *2 /( y * + D3 ) 2 , a33 = 0. Note that a12 < 0, a 22 > 0, a 23 < 0, and a32 > 0. By Routh-
Hurwitz criterion, the equilibrium point is asymptotically stable if the characteristic
equation has negative roots, or has one negative root and a complex pair with negative real
parts, that is if s1 > 0, s2 > 0, s3 > 0, and s1s2 − s3 > 0. For s3 > 0, we require a11 < 0. For
s1 > 0, we require a11 + a 22 < 0. Now, s1s2 − s3 = ( a11 + a22 )( a12 a 21 − a11a 22 ) +
a 22 a 23a32 . The required conditions are obtained from s2 > 0, a11 < 0, a11 + a 22 < 0, and
s1s2 − s3 > 0. For the given set of parameter values, we get the characteristic equation as
λ3 + 1.05876λ2 − 0.041241λ + 0.148856 = 0. The equation has one negative root −1.197,
and a complex pair with positive real parts. The equilibrium point is unstable.
6. Denote T1 = ( x 2 / i ) + x + a , and T2 = ( y 2 / i ) + y + a. From the first equation, we get
( dx / dt ) ≤ x (a1 − b1 x ), that is x(t ) ≤ [a1 /(b1 + ke − a1t )], where k is a constant depending
on the initial condition. As t → ∞, x(t ) ≤ a1 / b1 . Equilibrium points of the system are
E0 (0, 0, 0), E1 ( a1 / b1 , 0, 0), E2 ( xˆ , yˆ , 0), where x̂ is the solution of the quadratic equation
a 2 x 2 + i ( a2 − w1 ) x + ia2 a = 0, and yˆ = (a1 − b1 xˆ )T1 ( xˆ ) / w, 0 < x̂ < ( a1 / b1 ). The positive
equilibrium point is E3 ( x* , y * , z * ) , where y * is a positive root of the equation

35
cy 2 + i( c − w3 ) y + cia = 0. We require c < w3 , and i( w3 − c ) 2 > 4ac 2 . x* is a positive root
of the equation f ( x) = b1 x 3 + (ib1 − a1 ) x 2 + i (b1a − a1 ) x + i ( wy * − a1a ) = 0. Now,
f (0) = i ( wy * − a1a ), and f ( a1 / b1 ) = iwy * > 0. If y * < a1a / w, then f (0) f ( a1 / b1 ) < 0,
and a positive root exists in ( 0, a1 / b1 ). z * is given by z* = [T2 ( y* ) / w2 ]
[ −a 2 + {w1 x* / T1 ( x* )}]. Since z * > 0, we require {w1 x* / T1 ( x * )} > a 2 . The conditions
for the existence of a positive equilibrium point are c < w3 , i( w3 − c) 2 > 4ac 2 , y * < a1a
/ w, and {w1 x* / T1 ( x * )} > a 2 .
At the equilibrium point E0 (0, 0, 0), the eigenvalues of the Jacobian matrix J are a1 ,
−c, − a 2 . E0 (0, 0, 0) is unstable. . It is a saddle point with a stable manifold in y-z direction
ad unstable manifold in x direction. At the equilibrium point E1 ( a1 / b1 , 0, 0), the
eigenvalues of the Jacobian matrix J are − c, − a1 , a 22 , where a 22 = −a 2 + [ w1a1 /{b1
T1 ( a1 / b1 )}]. The equilibrium point is asymptotically stable for {a 2 b1T1 ( a1 / b1 )} > w1a1. At
the equilibrium point E2 ( xˆ , yˆ , 0), the characteristic equation of J is ( a33 − λ )(λ2 −
a 11 λ − a12 a21 ) = 0. One eigenvalue is λ1 = a33 , which is negative when [ w3 yˆ / T2 ( yˆ )]
< c. Now, a12 = −[ wxˆ / T1 ( xˆ )] < 0 ; a 21 = [ w1 yˆ (ia − xˆ 2 )] /[iT12 ( xˆ )] > 0, when xˆ 2 < ia ; and
a 11 < 0, when wyˆ (i + 2 xˆ ) < b1iT12 ( xˆ ). By Routh-Hurwitz criterion, the eigenvalues λ2 , λ3
are negative or has negative real parts when the above three conditions are satisfied. The
equilibrium point E2 ( xˆ , yˆ , 0) is asymptotically stable. At E3 ( x* , y * , z * ) , the characteristic
equation of J is λ3 + s1λ2 + s2 λ + s3 = 0, where s1 = −(a11 + a 22 ), s2 = ( a11a 22 − a12 a 21
− a23a32 ), s3 = a11a 23a 32 , a11 = x * [− b1 + {wy * ( 2 x * + i ) /(iT12 ( x * )}], a12 = wx * / T1
( x * ), a13 = 0, a 21 = [ w1 y * (ai − x *2 ) /{iT12 ( x * )}], a 22 = w2 y * z * ( 2 y * + i ) /{iT22 ( y * )}, a 23 =
= − w2 y * / T2 ( y * ), a31 = 0, a32 = w3 z * ( ai − y *2 ) / T22 ( y * ), a 33 = 0. Note that a12 < 0,
a22 > 0, and a 23 < 0. By Routh-Hurwitz criterion, the equilibrium point is asymptotically
stable if the characteristic equation has negative roots, or has one negative root and a
complex pair with negative real parts, that is if s1 > 0, s2 > 0, s3 > 0, and s1s2 − s3 > 0,
(sufficient conditions). For s3 > 0, we require a11 < 0. For s1 > 0, we require
a11 + a 22 < 0. Now, s1s2 − s3 = ( a11 + a 22 )( a12 a 21 − a11a22 ) + a22a23a32 . The required
conditions are obtained from s2 > 0, a11 < 0, a11 + a 22 < 0, and s1s2 − s3 > 0. Consider
the given set of parameter values. y * is the positive solution of the equation
cy 2 + i( c − w3 ) y + cia = 0, that is, 0.25 y 2 − 0.405 y + 0.075 = 0. The roots are
y1* = 0.213259 , y *2 = 1.406741 . We find that corresponding to y 2* there is no positive root
for the cubic equation for x . For *
y1* , *
the cubic equation in x has a positive root x1* = 0.9
11071, and a complex pair with negative real part. We obtain z * = 0.329525. For this
value of the equilibrium point, we get the characteristic equation as λ3 + 0.528227λ2 −

36
 0.107421λ + 0.084325 = 0. The equation has one negative root − 0.796175, and a complex
pair with positive real parts. The equilibrium point is unstable.
7. Denote T1 = ( x 2 / i1 ) + x + a, and T2 = ( y 2 / i2 ) + y + a. From the first equation, we obtain
0 < x < ( a1 / b1 ). Equilibrium points of the system are E0 (0, 0, 0), E1 (a1 / b1 , 0, 0),
E2 ( x , y , 0), where ~ x is the solution of the a 2 x 2 + i1 ( a2 − w1 ) x + aa2 i1 = 0, and
~ ~
y = ( a1 − b1 ~ x )T1 / w, 0 < x < a1 b1 . The positive equilibrium point is E3 ( x* , y* , z * ), where
y * = ( w3 − cd ) / c, w3 > cd . x* is a positive root of the equation f ( x ) = b1 x 3 + (i1b1 − a1 )
x 2 + i1 ( ab1 − a1 ) x + [i1 ( wy * − aa1 )] = 0, 0 < x* < (a1 / b1 ). Now, f (0) = [i1 ( wy * − aa1 )] < 0,
when y * < aa1 / w; and f ( a1 / b1 ) = i1wy * > 0. Hence, there exists a positive root in
( 0, a1 / b1 ). From the second equation, we get z * = [{w1 x * − a 2T1 ( x * )}T2 ( y * )] /{w2T1 ( x * )}.
Since z * > 0, we require w1 x* > a 2T1 ( x* ). In summary, the conditions for the existence of
a positive equilibrium point are cd < w3 , y * < a1a / w, and w1 x * > a 2T1 ( x * ). At the
equilibrium point E0 (0, 0, 0), the eigenvalues of the Jacobian matrix are a1 , − a2 , 0.
E0 (0, 0, 0) is a non-hyperbolic fixed point. Furthermore, as one eigenvalue is positive, and
another one is negative, E0 is always non-stable. At the equilibrium point E1 ( a1 / b1 , 0, 0),
the eigenvalues of the Jacobian matrix J are 0, − a1 , a22 , where a22 = −a2 + [ w1a1 /{b1T1
( a1 / b1 )}]. The equilibrium point is also a non-hyperbolic fixed point. At the equilibrium
point E2 = ( x , y ,0) , one of the eigenvalues is zero and the other eigenvalues are roots of
λ2 − a11λ − a12 a 21 = 0. If a11 < 0, and a12 a21 < 0, then both the eigenvalues are negative
or a complex pair with negative real parts. If a11 > 0, and a12 a 21 > 0, then one of the
eigenvalues is positive. E 2 is a non-hyperbolic point. The dynamical behavior near these
equilibrium points can be stable, periodic or even chaotic. At E3 ( x* , y * , z * ) , the
characteristic equation of J is λ3 + s1λ2 + s2 λ + s3 = 0, where s1 = −(a11 + a 22 ),
s2 = ( a11a 22 − a12 a 21 − a 23a 32 ), s3 = a11a 23a 32 , a11 = x*[ −b1 + {wy * ( 2 x* + i1 ) /(i1T12
( x * )}], a12 = − wx * / T1 ( x * ), a13 = 0, a 21 = [ w1 y * ( ai1 − x *2 ) /{i1T12 ( x * )}], a 22 = w2
y * z* ( 2 y * + i2 ) /{i2T22 ( y* )}, a31 = 0, a23 = − w2 y * / T2 ( y * ), a32 = c 2 z * / w3 , a33 = 0.
Note that a12 < 0, a 22 > 0, a 23 < 0, and a32 > 0. By Routh-Hurwitz criterion, the
equilibrium point is asymptotically stable if the characteristic equation has negative roots,
or has one negative root and a complex pair with negative real parts, that is if
s1 > 0, s2 > 0, s3 > 0, and s1s2 − s3 > 0, (sufficient conditions). For s3 > 0, we require
a11 < 0. For s1 > 0, we require a11 + a 22 < 0. Both the conditions are satisfied if
[ wx * ( 2 x * + i1 ) /(i1T12 ( x * )] + [ w2 z * ( 2 y * + i2 ) /{i2T22 ( y * )} < (b1 x * / y * ). Now, s1s2 − s3 =
( a11 + a 22 )( a12 a 21 − a11a 22 ) + a 22 a 23a 32 . The required conditions are the above
condition, s2 > 0, and s1s2 − s3 > 0. Consider the given set of parameter values.
(i) We obtain, y * = 625 / 27. x* is a positive root of the equation 0.075x 3 + 1.125x 2 −
73.575 x + 211.416666 = 0. We get two positive roots x1* = 22.718741, x2* = 3.043907.
37
 Corresponding to x1* , we get z * = 70.790948. Corresponding to x2* , we get a negative
value for z * . Therefore, the positive equilibrium point is (22.718741, 23.148148, 70.790
948). We get the characteristic equation as λ3 + 1.03824λ2 − 0.126348λ + 0.000776 = 0.
The equation has one negative root − 1.14881, and the other roots are 0.104077, 0.00649.
The equilibrium point is unstable.
(ii) We obtain, y * = 330 / 13. x* is a positive root of the equation 0.088 x 3 + 1.71x 2 −
68.778 x + 312.057692 = 0. We get two positive roots x1* = 15.812703, x*2 = 5.503507.
Corresponding to x1* , we obtain z * = 22.974668. Corresponding to x *2 , we find z * < 0.
Therefore, we get one positive equilibrium point (15.812703, 25.384615, 22.974668). We
get the characteristic equation as λ3 + 0.532106λ2 + 0.027195λ + 0.000417 = 0. The
equation has one negative root − 0.47692, and the complex pair −0.027594 ± 0.01064i.
The equilibrium point is asymptotically stable.
The chaotic attractor for the first set of parameter values is given in Fig. 4.6a. Stable limit
cycles for the second set of parameter values are given in Fig. 4.6b.
8 The model is numerically integrated to get the time series using MATLAB 7.5. Since every
non-linear system has a finite amount of transients, the data points representing transient
behavior were discarded. Phase portraits were drawn using this data to obtain the geometry
of the attractors. The chaotic attractor and its corresponding time series are depicted in
Figs.4.7a-b. We note that the solution of the model system is very sensitive to the death rate
parameter c in the range 0.3 ≤ c ≤ 0.318 . The route of chaos is through the cascade of
period doubling for decreasing values of the control parameter c. Other important changes
in the chaotic set include an interior crisis in which a chaotic attractor undergoes a sudden
increase in the size of the attractor. Closed curves marked as ‘A’ and ‘B’ in Fig.4.8
correspond to invariant KAM tori in the phase space. When the bifurcation parameter c is
decreased further, these curves break and give rise to chaotic dynamics.

(a) (b)
Fig.4.6a. Chaotic attractor the first set of parameter values. (b) Stable limit cycles for the
second set of parameter values. (From Upadhyay, R. K., Rai, V., Raw, S. N. Challenges of
living in the harsh environments: A mathematical modelling study. Appl. Math. Comp., 217,
10105–10117. Copyright 2011, Elsevier. Reprinted with permission).

38
Fig.4.7.(a) Chaotic attractor with initial condition [0.8848, 0.2903, 0.0002]. (b) Temporal
evolution of population densities x(t ), y (t ) and z (t ) for the chaotic attractor.
(From Upadhyay, R. K., Banerjee, M., Parshad, R. D., Raw, S. N. Deterministic chaos versus
stochastic oscillation in a prey-predator-top predator model. Math. Model. Anal.. 16(3), 343–364.
Copyright 2011, Taylor & Francis. Reprinted with permission).

Fig.4.8. Bifurcation diagram of the model system with c as a control parameter.

(From Upadhyay, R. K., Banerjee, M., Parshad, R. D., Raw, S. N. Deterministic chaos versus
stochastic oscillation in a prey-predator-top predator model. Math. Model. Anal.. 16(3), 343–364.
Copyright 2011, Taylor & Francis. Reprinted with permission).

39
 K22377
ISBN: 978-1-4822-2453-5
90000

w w w. c r c p r e s s . c o m
9 781482 224535
Another random document with
no related content on Scribd:
we were allowed to marry any of them, and were only precluded
from sexual intercourse with them without marriage, most people
would constantly have become guilty of misconduct with them. But
as they are entirely forbidden to us, and sexual intercourse with
them is most emphatically denounced unto us as a capital crime, or
a sin punishable with extinction (karet), and as there is no means of
ever legalizing such intercourse, there is reason to expect that
people will not seek it, and will not think of it. That the persons
included in that prohibition are, as we have stated, at hand and
easily accessible, is evident. For as a rule, the mother of the wife,
the grandmother, the daughter, the granddaughter, and the sister-in-
law, are mostly with her; the husband meets them always when he
goes out, when he comes in, and when he is at his work. The wife
stays also frequently in the house of her husband’s brother, father, or
son. It is also well known that we are often in the company of our
sisters, our aunts, and the wife of our uncle, and are frequently
brought up together with them. These are all the relatives which we
must not marry. This is one of the reasons why intermarriage with a
near relative is forbidden. But according to my opinion the
prohibition serves another object, namely, to inculcate chastity into
our hearts. Licence between the root and the branch, between a
man and his mother, or his daughter, is outrageous. The intercourse
between root and branch is forbidden, and it makes no difference
whether the male element is the root or the branch, or both root
and branch combine in the intercourse with a third person, so that
the same individual cohabits with the root and with the branch. On
this account it is prohibited to marry a woman and her mother, the
wife of the father or of the son; for in all these cases there is the
intercourse between one and the same person on the one side and
root and branch on the other.

The law concerning brothers is like the law concerning root and
branch. The sister is forbidden, and so is also the sister of the wife
and the wife of the brother; because in the latter cases two persons
who are considered like root and branch, cohabit with the same
person. But in these prohibitions brothers and sisters are partly
considered as root and branch and partly as one body; the sister of
the mother is therefore like the mother, and the [377]sister of the
father like the father, and both are prohibited; and since the
daughter of the parent’s brother or sister is not included in the
number of prohibited relatives, so may we also marry the daughter
of the brother or the sister. The apparent anomaly, that the brother
of the father may marry a woman that has been the wife of his
brother’s son, whilst the nephew must not marry a woman that has
been the wife of his father’s brother, can be explained according to
the above-mentioned first reason. For the nephew is frequently in
the house of his uncle, and his conduct towards the wife of his uncle
is the same as that towards his brother’s wife. The uncle, however, is
not so frequent in the house of his nephew, and he is consequently
less intimate with the wife of his nephew; whilst in the case of father
and son, the familiarity of the father with his daughter-in-law is the
same at that of the son with the wife of his father, and therefore the
law and punishment is the same for both [father and son]. The
reason why it is prohibited to cohabit with a menstruous woman
(Lev. xviii. 19) or with another man’s wife (ibid. 20), is obvious, and
requires no further explanation.

It is well known that we must not indulge in any sensual enjoyment

whatever with the persons included in the above prohibitions; we
must not even look at them if we intend to derive pleasure
therefrom. We have explained this in “the laws about forbidden
sexual intercourse” (Hilkot issure bïah, xxi. 1–2), and shown that
according to the Law we must not even engage our thoughts with
the act of cohabitation (ibid. 19) or irritate the organ of generation;
and when we find ourselves unintentionally in a state of irritation,
we must turn our mind to other thoughts, and reflect on some other
thing till we are relieved. Our Sages (B.T. Kidd 30b), in their moral
lessons, which give perfection to the virtuous, say as follows: “My
son, if that monster meets you, drag it to the house of study. It will
melt if it is of iron; it will break in pieces if it is of stone: as is said in
Scripture, ‘Is not my word like a fire? saith the Lord, and like a
hammer that breaketh the rock in pieces?’ ” (Jer. xxiii. 29). The
author of this saying thus exhorts his son to go to the house of
study when he finds his organ of generation in an irritated state. By
reading, disputing, asking, and listening to questions, the irritation
will certainly cease. See how properly the term monster is employed,
for that irritation is indeed like a monster. Not only religion teaches
this lesson, the philosophers teach the same. I have already quoted
verbatim the words of Aristotle. He says: “The sense of touch which
is a disgrace to us, leads us to indulge in eating and sensuality,” etc.
He calls people degraded who seek carnal pleasures and devote
themselves to gastronomy; he denounces in extenso their low and
objectionable conduct, and ridicules them. This passage occurs in his
Ethics and in his Rhetoric.

In accordance with this excellent principle, which we ought strictly to

follow, our Sages teach us that we ought not to look at beasts or
birds in the moment of their copulation. According to my opinion,
this is the reason why the cross-breeding of cattle is prohibited (Lev.
xix. 19). It is a fact that animals of different species do not copulate
together, unless by force. It is well known that the low class of
breeders of mules are regularly engaged in this work. Our Law
objected to it that any Israelite should degrade himself by doing
these things, which require so much vulgarity and indecency, and
doing that which religion forbids us even to mention, how [378]much
more to witness or to practise, except when necessary. Cross-
breeding, however, is not necessary. I think that the prohibition to
bring together two species in any kind of work, as included in the
words, “Thou shalt not plow with an ox and an ass together” (Deut.
xxii. 10), is only a preventive against the intercourse of two species.
For if it were allowed to join such together in any work, we might
sometimes also cause their intercourse. That this is the reason of the
commandment is proved by the fact that it applies to other animals
besides ox and ass; it is prohibited to plow not only with ox and ass
together, but with any two kinds. But Scripture mentions as an
instance that which is of regular occurrence.

As regards circumcision, I think that one of its objects is to limit

sexual intercourse, and to weaken the organ of generation as far as
possible, and thus cause man to be moderate. Some people believe
that circumcision is to remove a defect in man’s formation; but every
one can easily reply: How can products of nature be deficient so as
to require external completion, especially as the use of the fore-skin
to that organ is evident. This commandment has not been enjoined
as a complement to a deficient physical creation, but as a means for
perfecting man’s moral shortcomings. The bodily injury caused to
that organ is exactly that which is desired; it does not interrupt any
vital function, nor does it destroy the power of generation.
Circumcision simply counteracts excessive lust; for there is no doubt
that circumcision weakens the power of sexual excitement, and
sometimes lessens the natural enjoyment; the organ necessarily
becomes weak when it loses blood and is deprived of its covering
from the beginning. Our Sages (Beresh. Rabba, c. 80) say distinctly:
It is hard for a woman, with whom an uncircumcised had sexual
intercourse, to separate from him. This is, as I believe, the best
reason for the commandment concerning circumcision. And who was
the first to perform this commandment? Abraham, our father! of
whom it is well known how he feared sin; it is described by our
Sages in reference to the words, “Behold, now I know that thou art
a fair woman to look upon” (Gen. xii. 11).
There is, however, another important object in this commandment.
It gives to all members of the same faith, i.e., to all believers in the
Unity of God, a common bodily sign, so that it is impossible for any
one that is a stranger, to say that he belongs to them. For
sometimes people say so for the purpose of obtaining some
advantage, or in order to make some attack upon the Jews. No one,
however, should circumcise himself or his son for any other reason
but pure faith; for circumcision is not like an incision on the leg, or a
burning in the arm, but a very difficult operation. It is also a fact that
there is much mutual love and assistance among people that are
united by the same sign when they consider it as [the symbol of] a
covenant. Circumcision is likewise the [symbol of the] covenant
which Abraham made in connexion with the belief in God’s Unity. So
also every one that is circumcised enters the covenant of Abraham
to believe in the unity of God, in accordance with the words of the
Law, “To be a God unto thee, and to thy seed after thee” (Gen. xvii.
7). This purpose of the circumcision is as important as the first, and
perhaps more important.

This law can only be kept and perpetuated in its perfection, if

circumcision is performed when the child is very young, and this for
three good reasons. First, if the operation were postponed till the
boy had grown up, [379]he would perhaps not submit to it. Secondly,
the young child has not much pain, because the skin is tender, and
the imagination weak; for grown-up persons are in dread and fear of
things which they imagine as coming, some time before these
actually occur. Thirdly, when a child is very young, the parents do
not think much of him; because the image of the child, that leads
the parents to love him, has not yet taken a firm root in their minds.
That image becomes stronger by the continual sight; it grows with
the development of the child, and later on the image begins again to
decrease and to vanish. The parents’ love for a new-born child is not
so great as it is when the child is one year old; and when one year
old, it is less loved by them than when six years old. The feeling and
love of the father for the child would have led him to neglect the law
if he were allowed to wait two or three years, whilst shortly after
birth the image is very weak in the mind of the parent, especially of
the father who is responsible for the execution of this
commandment. The circumcision must take place on the eighth day
(Lev. xii. 3), because all living beings are after birth, within the first
seven days, very weak and exceedingly tender, as if they were still in
the womb of their mother; not until the eighth day can they be
counted among those that enjoy the light of the world. That this is
also the case with beasts may be inferred from the words of
Scripture: “Seven days shall it be under the dam” (Lev. xxii. 27), as if
it had no vitality before the end of that period. In the same manner
man is circumcised after the completion of seven days. The period
has been fixed, and has not been left to everybody’s judgment.

The precepts of this class include also the lesson that we must not
injure in any way the organs of generation in living beings (ibid. xxii.
24). The lesson is based on the principle of “righteous statutes and
judgment” (Deut. iv. 8); we must keep in everything the golden
mean; we must not be excessive in love, but must not suppress it
entirely; for the Law commands, “Be fruitful, and multiply” (Gen. i.
22). The organ is weakened by circumcision, but not destroyed by
the operation. The natural faculty is left in full force, but is guarded
against excess. It it prohibited for an Israelite “that is wounded in
the stones, or hath his privy member cut off” (Deut. xxiii. 2), to
marry an Israelitish woman; because the sexual intercourse is of no
use and of no purpose; and that marriage would be a source of ruin
to her, and to him who would claim her. This is very clear.

In order to create a horror of illicit marriages, a bastard was not

allowed to marry an Israelitish woman (ibid. xxiii. 3); the adulterer
and the adulteress were thus taught that by their act they bring
upon their seed irreparable injury. In every language and in every
nation the issue of licentious conduct has a bad name; the Law
therefore raises the name of the Israelites by keeping them free
from the admixture of bastards. The priests, who have a higher
sanctity, are not allowed to marry a harlot, or a woman that is
divorced from her husband, or that is profane (Lev. xii. 7); the high-
priest, the noblest of the priests, must not marry even a widow, or a
woman that has had sexual intercourse of any kind (ibid. xii. 14). Of
all these laws the reason is obvious. If bastards were prohibited to
marry any member of the congregation of the Lord, how much more
rigidly had slaves and handmaids to be excluded. The reason of the
prohibition of intermarriage [380]with other nations is stated in the
Law: “And thou take of their daughters unto thy sons, and their
daughters go a whoring after their gods, and make thy sons go a
whoring after their gods” (Exod. xxxiv. 16).

Most of the “statutes” (ḥukkim), the reason of which is unknown to

us, serve as a fence against idolatry. That I cannot explain some
details of the above laws or show their use is owing to the fact that
what we hear from others is not so clear as that which we see with
our own eyes. Thus my knowledge of the Sabean doctrines, which I
derived from books, is not as complete as the knowledge of those
who have witnessed the public practice of those idolatrous customs,
especially as they have been out of practice and entirely extinct
since two thousand years. If we knew all the particulars of the
Sabean worship, and were informed of all the details of those
doctrines, we would clearly see the reason and wisdom of every
detail in the sacrificial service, in the laws concerning things that are
unclean, and in other laws, the object of which I am unable to state.
I have no doubt that all these laws served to blot out wrong
principles from man’s heart, and to exterminate the practices which
are useless, and merely a waste of time in vain and purposeless
things. Those principles have turned the mind of the people away
from intellectual research and useful actions. Our prophets therefore
describe the ways of the idolaters as follows: “(They go) after vain
things which cannot profit nor deliver; for they are vain” (1 Sam. xii.
21); “Surely our fathers have inherited lies, vanity and things
wherein there is no profit” (Jer. xvi. 19). Consider how great the evil
consequences of idolatry are, and say whether we ought with all our
power to oppose it or not! Most of the precepts serve, as has been
stated by us, as a mere fence against those doctrines [of idolatry],
and relieve man from the great and heavy burdens, from the pains
and inflictions which formed part of the worship of idols. Every
positive or negative precept, the reason of which is unknown to
thee, take as a remedy against some of those diseases with which
we are unacquainted at present, thank God. This should be the
belief of educated men who know the true meaning of the following
divine dictum: “I said not unto the seed of Jacob, Seek me in vain”
(Isa. xlv. 19).

I have now mentioned all the commandments of these fourteen

classes one by one, and pointed out the reason of each of them,
with the exception of a few for which I was unable to give the
reason, and of some details of less importance; but implicitly we
have given the reason even of these, and every intelligent reader will
easily find it.

The reasons of the Precepts are now complete.

[Contents]
 CHAPTER L
There are in the Law portions which include deep wisdom, but have
been misunderstood by many persons; they require, therefore, an
explanation. I mean the narratives contained in the Law which many
consider as being of no use whatever; e.g., the list of the various
families descended from Noah, with their names and their territories
(Gen. x.); the sons of Seir the Horite (ibid. xxxvi. 20–30); the kings
that reigned in Edom (ibid. 31, seq.); and the like. There is a saying
of our Sages (B.T. Sanh. 99b) that the wicked king Manasse
frequently held disgraceful meetings for the sole purpose of
criticising such passages of the Law. “He held meetings and made
blasphemous [381]observations on Scripture, saying, Had Moses
nothing else to write than, And the sister of Lotan was Timna” (Gen.
xxxvi. 22)? With reference to such passages, I will first give a
general principle, and then discuss them seriatim, as I have done in
the exposition of the reasons of the precepts.

Every narrative in the Law serves a certain purpose in connexion

with religious teaching. It either helps to establish a principle of
faith, or to regulate our actions, and to prevent wrong and injustice
among men; and I will show this in each case.

It is one of the fundamental principles of the Law that the Universe

has been created ex nihilo, and that of the human race, one
individual being, Adam, was created. As the time which elapsed from
Adam to Moses was not more than about two thousand five hundred
years, people would have doubted the truth of that statement if no
other information had been added, seeing that the human race was
spread over all parts of the earth in different families and with
different languages, very unlike the one to the other. In order to
remove this doubt the Law gives the genealogy of the nations (Gen.
v. and x.), and the manner how they branched off from a common
root. It names those of them who were well known, and tells who
their fathers were, how long and where they lived. It describes also
the cause that led to the dispersion of men over all parts of the
earth, and to the formation of their different languages, after they
had lived for a long time in one place, and spoken one language
(ibid. xi.), as would be natural for descendants of one person. The
accounts of the flood (ibid. vi.-viii.) and of the destruction of Sodom
and Gomorrah (ibid. xix.), serve as an illustration of the doctrine that
“Verily there is a reward for the righteous; verily He is a God that
judgeth in the earth” (Ps. lviii. 12).

The narration of the war among the nine kings (ibid. xiv.) shows
how, by means of a miracle, Abraham, with a few undisciplined men,
defeated four mighty kings. It illustrates at the same time how
Abraham sympathized with his relative, who had been brought up in
the same faith, and how he exposed himself to the dangers of
warfare in order to save him. We further learn from this narrative
how contented and satisfied Abraham was, thinking little of property,
and very much of good deeds; he said, “I will not take from a thread
even to a shoe-latchet” (Gen. xiv. 23).

The list of the families of Seir and their genealogy is given in the
Law (ibid. xxxvi. 20–30), because of one particular commandment.
For God distinctly commanded the Israelites concerning Amalek to
blot out his name (Deut. xxv. 17–19). Amalek was the son of Eliphas
and Timna, the sister of Lotan (ibid. xxxvi. 12). The other sons of
Esau were not included in this commandment. But Esau was by
marriage connected with the Seirites, as is distinctly stated in
Scripture; and Seirites were therefore his children; he reigned over
them; his seed was mixed with the seed of Seir, and ultimately all
the countries and families of Seir were called after the sons of Esau
who were the predominant family, and they assumed more
particularly the name Amalekites, because these were the strongest
in that family. If the genealogy of these families of Seir had not been
described in full they would all have been killed, contrary to the plain
words of the commandment. For this reason the Seirite families are
fully described, as if to say, the people that live in Seir and in the
kingdom of Amalek are not all Amalekites; they [382]are the
descendants of some other man, and are called Amalekites because
the mother of Amalek was of their tribe. The justice of God thus
prevented the destruction of an [innocent] people that lived in the
midst of another people [doomed to extirpation]; for the decree was
only pronounced against the seed of Amalek. The reason of this
decree has already been stated by us (p. 205).

The kings that have reigned in the land of Edom are enumerated
(Gen xxxvi. 31, seq.) on account of the law, “Thou mayst not set a
stranger over thee, which is not thy brother” (Deut. xvii. 15). For of
these kings none was an Edomite; wherefore each king is described
by his native land; one king from this place, another king from that
place. Now I think that it was then well known how these kings that
reigned in Edom conducted themselves, what they did, and how
they humiliated and oppressed the sons of Esau. Thus God reminded
the Israelites of the fate of the Edomites, as if saying unto them,
Look unto your brothers, the sons of Esau, whose kings were so and
so, and whose deeds are well known. [Learn therefrom] that no
nation ever chose a foreigner as king without inflicting thereby some
great or small injury upon the country. In short, what I remarked in
reference to our ignorance of the Sabean worship, applies also to
the history of those days. If the religious rules of the Sabeans and
the events of those days were known to us, we should be able to
see plainly the reason for most of the things mentioned in the
Pentateuch.
It is also necessary to note the following observations. The view we
take of things described by others is different from the view we take
of things seen by us as eye-witnesses. For that which we see
contains many details which are essential, and must be fully
described. The reader of the description believes that it contains
superfluous matter, or useless repetition, but if he had witnessed the
event of which he reads, he would see the necessity of every part of
the description. When we therefore notice narratives in the Torah,
which are in no connexion with any of the commandments, we are
inclined to think that they are entirely superfluous, or too lengthy, or
contain repetitions; but this is only because we do not see the
particular incidents which make those narratives noteworthy. Of this
kind is the enumeration of the stations [of the Israelites in the
wilderness] (Num. xxxiii.). At first sight it appears to be entirely
useless; but in order to obviate such a notion Scripture says, “And
Moses wrote their goings out according to their journeys by the
commandment of the Lord” (ibid. ver. 2). It was indeed most
necessary that these should be written. For miracles are only
convincing to those who witnessed them; whilst coming generations,
who know them only from the account given by others, may
consider them as untrue. But miracles cannot continue and last for
all generations; it is even inconceivable [that they should be
permanent]. Now the greatest of the miracles described in the Law
is the stay of the Israelites in the wilderness for forty years, with a
daily supply of manna. This wilderness, as described in Scripture,
consisted of places “wherein were fiery serpents and scorpions, and
drought, where there was no water” (Deut. viii. 15); places very
remote from cultivated land, and naturally not adapted for the
habitation of man, “It is no place of seed, or of figs, or of vines, or
of pomegranates, neither is there any water to drink” (Num. xx. 5);
“A land that no man passed through, and where no man dwelt” (Jer.
ii. 6). [In reference to the stay of [383]the Israelites in the
wilderness], Scripture relates, “Ye have not eaten bread, neither
have ye drunk wine or strong drink” (Deut. xix. 5). All these miracles
were wonderful, public, and witnessed by the people. But God knew
that in future people might doubt the correctness of the account of
these miracles, in the same manner as they doubt the accuracy of
other narratives; they might think that the Israelites stayed in the
wilderness in a place not far from inhabited land, where it was
possible for man to live [in the ordinary way]; that it was like those
deserts in which Arabs live at present; or that they dwelt in such
places in which they could plow, sow, and reap, or live on some
vegetable that was growing there; or that manna came always down
in those places as an ordinary natural product; or that there were
wells of water in those places. In order to remove all these doubts
and to firmly establish the accuracy of the account of these miracles,
Scripture enumerates all the stations, so that coming generations
may see them, and learn the greatness of the miracle which enabled
human beings to live in those places forty years.

For this very reason Joshua cursed him who would ever build up
Jericho (Josh. vi. 26); the effect of the miracle was to remain for
ever, so that any one who would see the wall sunk in the ground
would understand that it was not in the condition of a building pulled
down by human hands, but sunk through a miracle. In a similar
manner the words, “At the commandment of the Lord the children of
Israel journeyed, and at the commandment of the Lord they pitched”
(Num. ix. 18), would suffice as a simple statement of facts; and the
reader might at first sight consider as unnecessary additions all the
details which follow, viz., “And when the cloud tarried long.… And so
it was when the cloud was a few days.… Or whether it were two
days,” etc. (ibid. ix. 19–22). But I will show you the reason why all
these details are added. For they serve to confirm the account, and
to contradict the opinion of the nations, both of ancient and modern
times, that the Israelites lost their way, and did not know where to
go; that “they were entangled in the land” (Exod. xiv. 3); wherefore
the Arabs unto this day call that desert Al-tih, “the desert of going
astray,” imagining that the Israelites erred about, and did not know
the way. Scripture, therefore, clearly states and emphatically
declares that it was by God’s command that the journeyings were
irregular, that the Israelites returned to the same places several
times, and that the duration of the stay was different in each
station; whilst the stay in one place continued for eighteen years, in
another place it lasted one day, and in another one night. There was
no going astray, but the journey was regulated by “the rising of the
pillar of cloud” (Num. ix. 17). Therefore all these details are given.
Scripture clearly states that the way was near, known, and in good
condition; I mean the way from Horeb, whither they came
intentionally, according to the command of God, “Ye shall serve God
upon this mountain” (Exod. ii. 12), to Kadesh-barnea, the beginning
of inhabited land, as Scripture says, “Behold, we are now in Kadesh,
a city in the uttermost of thy border” (Num. xx. 16). That way was a
journey of eleven days; comp. “Eleven days’ journey from Horeb, by
the way of mount Seir, unto Kadesh-barnea” (Deut. i. 3). In such a
journey it is impossible to err about for forty years; but Scripture
states the cause of the delay. [384]

In like manner there is a good reason for every passage the object
of which we cannot see. We must always apply the words of our
Sages: “It is not a vain thing for you” (Deut. xxxii. 47), and if it
seems vain, it seems your fault.

[Contents]
 CHAPTER LI
The present chapter does not contain any additional matter that has
not been treated in the [previous] chapters of this treatise. It is a
kind of conclusion, and at the same time it will explain in what
manner those worship God who have obtained a true knowledge
concerning God; it will direct them how to come to that worship,
which is the highest aim man can attain, and show how God
protects them in this world till they are removed to eternal life.

I will begin the subject of this chapter with a simile. A king is in his
palace, and all his subjects are partly in the country, and partly
abroad. Of the former, some have their backs turned towards the
king’s palace, and their faces in another direction; and some are
desirous and zealous to go to the palace, seeking “to inquire in his
temple,” and to minister before him, but have not yet seen even the
face of the wall of the house. Of those that desire to go to the
palace, some reach it, and go round about in search of the entrance
gate; others have passed through the gate, and walk about in the
ante-chamber; and others have succeeded in entering into the inner
part of the palace, and being in the same room with the king in the
royal palace. But even the latter do not immediately on entering the
palace see the king, or speak to him; for, after having entered the
inner part of the palace, another effort is required before they can
stand before the king—at a distance, or close by—hear his words, or
speak to him. I will now explain the simile which I have made. The
people who are abroad are all those that have no religion, neither
one based on speculation nor one received by tradition. Such are the
extreme Turks that wander about in the north, the Kushites who live
in the south, and those in our country who are like these. I consider
these as irrational beings, and not as human beings; they are below
mankind, but above monkeys, since they have the form and shape
of man, and a mental faculty above that of the monkey.

Those who are in the country, but have their backs turned towards
the king’s palace, are those who possess religion, belief, and
thought, but happen to hold false doctrines, which they either
adopted in consequence of great mistakes made in their own
speculations, or received from others who misled them. Because of
these doctrines they recede more and more from the royal palace
the more they seem to proceed. These are worse than the first
class, and under certain circumstances it may become necessary to
slay them, and to extirpate their doctrines, in order that others
should not be misled.

Those who desire to arrive at the palace, and to enter it, but have
never yet seen it, are the mass of religious people; the multitude
that observe the divine commandments, but are ignorant. Those
who arrive at the palace, but go round about it, are those who
devote themselves exclusively to the study of the practical law; they
believe traditionally in true principles of faith, and learn the practical
worship of God, but are not trained in philosophical treatment of the
principles of the Law, and do not endeavour to [385]establish the
truth of their faith by proof. Those who undertake to investigate the
principles of religion, have come into the ante-chamber; and there is
no doubt that these can also be divided into different grades. But
those who have succeeded in finding a proof for everything that can
be proved, who have a true knowledge of God, so far as a true
knowledge can be attained, and are near the truth, wherever an
approach to the truth is possible, they have reached the goal, and
are in the palace in which the king lives.

My son, so long as you are engaged in studying the Mathematical

Sciences and Logic, you belong to those who go round about the
palace in search of the gate. Thus our Sages figuratively use the
phrase: “Ben-zoma is still outside.” When you understand Physics,
you have entered the hall; and when, after completing the study of
Natural Philosophy, you master Metaphysics, you have entered the
innermost court, and are with the king in the same palace. You have
attained the degree of the wise men, who include men of different
grades of perfection. There are some who direct all their mind
toward the attainment of perfection in Metaphysics, devote
themselves entirely to God, exclude from their thought every other
thing, and employ all their intellectual faculties in the study of the
Universe, in order to derive therefrom a proof for the existence of
God, and to learn in every possible way how God rules all things;
they form the class of those who have entered the palace, namely,
the class of prophets. One of these has attained so much
knowledge, and has concentrated his thoughts to such an extent in
the idea of God, that it could be said of him, “And he was with the
Lord forty days,” etc. (Exod. xxxiv. 28); during that holy communion
he could ask Him, answer Him, speak to Him, and be addressed by
Him, enjoying beatitude in that which he had obtained to such a
degree that “he did neither eat bread nor drink water” (ibid.); his
intellectual energy was so predominant that all coarser functions of
the body, especially those connected with the sense of touch, were
in abeyance. Some prophets are only able to see, and of these some
approach near and see, whilst others see from a distance: comp.
“The Lord hath appeared from far unto me” (Jer. xxxi. 3). We have
already spoken of the various degrees of prophets; we will therefore
return to the subject of this chapter, and exhort those who have
attained a knowledge of God, to concentrate all their thoughts in
God. This is the worship peculiar to those who have acquired a
knowledge of the highest truths; and the more they reflect on Him,
and think of Him, the more are they engaged in His worship. Those,
however, who think of God, and frequently mention His name,
without any correct notion of Him, but merely following some
imagination, or some theory received from another person, are, in
my opinion, like those who remain outside the palace and distant
from it. They do not mention the name of God in truth, nor do they
reflect on it. That which they imagine and mention does not
correspond to any being in existence; it is a thing invented by their
imagination, as has been shown by us in our discussion on the
Divine Attributes (Part I. chap. 1.). The true worship of God is only
possible when correct notions of Him have previously been
conceived. When you have arrived by way of intellectual research at
a knowledge of God and His works, then commence to devote
yourselves to Him, try to approach Him and strengthen the intellect,
which is the link that joins you to Him. Thus Scripture says, “Unto
thee it was [386]showed, that thou mightest know that the Lord He is
God” (Deut. iv. 35); “Know therefore this day, and consider it in
thine heart, that the Lord He is God” (ibid. 36); “Know ye that the
Lord is God” (Ps. c. 3). Thus the Law distinctly states that the
highest kind of worship to which we refer in this chapter, is only
possible after the acquisition of the knowledge of God. For it is said,
“To love the Lord your God, and to serve Him with all your heart and
with all your soul” (Deut. xi. 13), and, as we have shown several
times, man’s love of God is identical with His knowledge of Him. The
Divine service enjoined in these words must, accordingly, be
preceded by the love of God. Our Sages have pointed out to us that
it is a service in the heart, which explanation I understand to mean
this: man concentrates all his thoughts on the First Intellect, and is
absorbed in these thoughts as much as possible. David therefore
commands his son Solomon these two things, and exhorts him
earnestly to do them: to acquire a true knowledge of God, and to be
earnest in His service after that knowledge has been acquired. For
he says, “And thou, Solomon my son, know thou the God of thy
father, and serve him with a perfect heart … if thou seek him, he will
be found of thee; but if thou forsake him, he will cast thee off for
ever” (1 Chron. xxviii. 9). The exhortation refers to the intellectual
conceptions, not to the imaginations; for the latter are not called
“knowledge,” but “that which cometh into your mind” (Ezek. xx. 32).
It has thus been shown that it must be man’s aim, after having
acquired the knowledge of God, to deliver himself up to Him, and to
have his heart constantly filled with longing after Him. He
accomplishes this generally by seclusion and retirement. Every pious
man should therefore seek retirement and seclusion, and should only
in case of necessity associate with others.

Note.—I have shown you that the intellect which emanates from
God unto us is the link that joins us to God. You have it in your
power to strengthen that bond, if you choose to do so, or to weaken
it gradually till it breaks, if you prefer this. It will only become strong
when you employ it in the love of God, and seek that love; it will be
weakened when you direct your thoughts to other things. You must
know that even if you were the wisest man in respect to the true
knowledge of God, you break the bond between you and God
whenever you turn entirely your thoughts to the necessary food or
any necessary business; you are then not with God, and He is not
with you; for that relation between you and Him is actually
interrupted in those moments. The pious were therefore particular to
restrict the time in which they could not meditate upon the name of
God, and cautioned others about it, saying, “Let not your minds be
vacant from reflections upon God.” In the same sense did David say,
“I have set the Lord always before me; because he is at my right
hand, I shall not be moved” (Ps. xvi. 8); i.e., I do not turn my
thoughts away from God; He is like my right hand, which I do not
forget even for a moment on account of the ease of its motions, and
therefore I shall not be moved, I shall not fall.

We must bear in mind that all such religious acts as reading the Law,
praying, and the performance of other precepts, serve exclusively as
the means of causing us to occupy and fill our mind with the
precepts of God, and free it from worldly business; for we are thus,
as it were, in communication with God, and undisturbed by any
other thing. If we, however, [387]pray with the motion of our lips,
and our face toward the wall, but at the same time think of our
business; if we read the Law with our tongue, whilst our heart is
occupied with the building of our house, and we do not think of
what we are reading; if we perform the commandments only with
our limbs, we are like those who are engaged in digging in the
ground, or hewing wood in the forest, without reflecting on the
nature of those acts, or by whom they are commanded, or what is
their object. We must not imagine that [in this way] we attain the
highest perfection; on the contrary, we are then like those in
reference to whom Scripture says, “Thou art near in their mouth,
and far from their reins” (Jer. xii. 2).

I will now commence to show you the way how to educate and train
yourselves in order to attain that great perfection.

The first thing you must do is this: Turn your thoughts away from
everything while you read Shemaʻ or during the Tefillah, and do not
content yourself with being devout when you read the first verse of
Shema, or the first paragraph of the prayer. When you have
successfully practised this for many years, try in reading the Law or
listening to it, to have all your heart and all your thought occupied
with understanding what you read or hear. After some time when
you have mastered this, accustom yourself to have your mind free
from all other thoughts when you read any portion of the other
books of the prophets, or when you say any blessing; and to have
your attention directed exclusively to the perception and the
understanding of what you utter. When you have succeeded in
properly performing these acts of divine service, and you have your
thought, during their performance, entirely abstracted from worldly
affairs, take then care that your thought be not disturbed by thinking
of your wants or of superfluous things. In short, think of worldly
matters when you eat, drink, bathe, talk with your wife and little
children, or when you converse with other people. These times,
which are frequent and long, I think, must suffice to you for
reflecting on everything that is necessary as regards business,
household, and health. But when you are engaged in the
performance of religious duties, have your mind exclusively directed
to what you are doing.

When you are alone by yourself, when you are awake on your couch,
be careful to meditate in such precious moments on nothing but the
intellectual worship of God, viz., to approach Him and to minister
before Him in the true manner which I have described to you—not in
hollow emotions. This I consider as the highest perfection wise men
can attain by the above training.

When we have acquired a true knowledge of God, and rejoice in that

knowledge in such a manner, that whilst speaking with others, or
attending to our bodily wants, our mind is all that time with God;
when we are with our heart constantly near God, even whilst our
body is in the society of men; when we are in that state which the
Song on the relation between God and man poetically describes in
the following words: “I sleep, but my heart waketh; it is the voice of
my beloved knocketh” (Song v. 2):—then we have attained not only
the height of ordinary prophets, but of Moses, our Teacher, of whom
Scripture relates: “And Moses alone shall come near before the Lord”
(ibid. xxxiv. 28); “But as for thee, stand thou here by me” (Deut. v.
28). The meaning of these verses has been explained by us. [388]

The Patriarchs likewise attained this degree of perfection; they

approached God in such a manner that with them the name of God
became known in the world. Thus we read in Scripture: “The God of
Abraham, the God of Isaac, and the God of Jacob.… This is My name
for ever” (Exod. iii. 15). Their mind was so identified with the
knowledge of God, that He made a lasting covenant with each of
them: “Then will I remember my covenant with Jacob,” etc. (Lev.
xxvi. 42). For it is known from statements made in Scripture that
these four, viz., the Patriarchs and Moses, had their minds exclusively
filled with the name of God, that is, with His knowledge and love;
and that in the same measure was Divine Providence attached to
them and their descendants. When we therefore find them also,
engaged in ruling others, in increasing their property, and
endeavouring to obtain possession of wealth and honour, we see in
this fact a proof that when they were occupied in these things, only
their bodily limbs were at work, whilst their heart and mind never
moved away from the name of God. I think these four reached that
high degree of perfection in their relation to God, and enjoyed the
continual presence of Divine Providence, even in their endeavours to
increase their property, feeding the flock, toiling in the field, or
managing the house, only because in all these things their end and
aim was to approach God as much as possible. It was the chief aim
of their whole life to create a people that should know and worship
God. Comp. “For I know him, that he will command his children and
his household after him” (Gen. xviii. 19). The object of all their
labours was to publish the Unity of God in the world, and to induce
people to love Him; and it was on this account that they succeeded
in reaching that high degree; for even those [worldly] affairs were
for them a perfect worship of God. But a person like myself must not
imagine that he is able to lead men up to this degree of perfection.
It is only the next degree to it that can be attained by means of the
above-mentioned training. And let us pray to God and beseech Him
that He clear and remove from our way everything that forms an
obstruction and a partition between us and Him, although most of
these obstacles are our own creation, as has several times been
shown in this treatise. Comp. “Your iniquities have separated
between you and your God” (Isa. lix. 2).
An excellent idea presents itself here to me, which may serve to
remove many doubts, and may help to solve many difficult problems
in metaphysics. We have already stated in the chapters which treat
of Divine Providence, that Providence watches over every rational
being according to the amount of intellect which that being
possesses. Those who are perfect in their perception of God, whose
mind is never separated from Him, enjoy always the influence of
Providence. But those who, perfect in their knowledge of God, turn
their mind sometimes away from God, enjoy the presence of Divine
Providence only when they meditate on God; when their thoughts
are engaged in other matters, divine Providence departs from them.
The absence of Providence in this case is not like its absence in the
case of those who do not reflect on God at all; it is in this case less
intense, because when a person perfect in his knowledge [of God] is
busy with worldly matters, he has not knowledge in actuality, but
only knowledge in potentiality [though ready to become actual]. This
person is then like a trained scribe when he is not writing. Those
who have no knowledge of God are like those who are [389]in
constant darkness and have never seen light. We have explained in
this sense the words: “The wicked shall be silent in darkness” (1
Sam. ii. 9) whilst those who possess the knowledge of God, and
have their thoughts entirely directed to that knowledge, are, as it
were, always in bright sunshine; and those who have the
knowledge, but are at times engaged in other themes, have then as
it were a cloudy day: the sun does not shine for them on account of
the cloud that intervenes between them and God.

Hence it appears to me that it is only in times of such neglect that

some of the ordinary evils befall a prophet or a perfect and pious
man; and the intensity of the evil is proportional to the duration of
those moments, or to the character of the things that thus occupy
their mind. Such being the case, the great difficulty is removed that
led philosophers to assert that Providence does not extend to every
individual, and that man is like any other living being in this respect,
viz., the argument based on the fact that good and pious men are
afflicted with great evils. We have thus explained this difficult
question even in accordance with the philosophers’ own principles.
Divine Providence is constantly watching over those who have
obtained that blessing which is prepared for those who endeavour to
obtain it. If man frees his thoughts from worldly matters, obtains a
knowledge of God in the right way, and rejoices in that knowledge, it
is impossible that any kind of evil should befall him while he is with
God, and God with him. When he does not meditate on God, when
he is separated from God, then God is also separated from him; then
he is exposed to any evil that might befall him; for it is only that
intellectual link with God that secures the presence of Providence
and protection from evil accidents. Hence it may occur that the
perfect man is at times not happy, whilst no evil befalls those who
are imperfect; in these cases what happens to them is due to
chance. This principle I find also expressed in the Law. Comp. “And I
will hide my face from them, and they shall be devoured, and many
evils and troubles shall befall them; so that they will say in that day,
Are not these evils come upon us, because our God is not among
us?” (Deut. xxxi. 17). It is clear that we ourselves are the cause of
this hiding of the face, and that the screen that separates us from
God is of our own creation. This is the meaning of the words: “And I
will surely hide my face in that day, for all the evils which they shall
have wrought” (ibid. ver. 18). There is undoubtedly no difference in
this regard between one single person and a whole community. It is
now clearly established that the cause of our being exposed to
chance, and abandoned to destruction like cattle, is to be found in
our separation from God. Those who have their God dwelling in their
hearts, are not touched by any evil whatever. For God says: “Fear
thou not, for I am with thee; be not dismayed, for I am thy God”
(Isa. xli. 10). “When thou passest through the waters, I will be with
thee; and through the rivers, they shall not overflow thee” (ibid. xliii.
2). For if we prepare ourselves, and attain the influence of the Divine
Intellect, Providence is joined to us, and we are guarded against all
evils. Comp. “The Lord is on my side; I will not fear; what can man
do unto me?” (Ps. cxviii. 6). “Acquaint now thyself with him, and be
at peace” (Job xxii. 21); i.e., turn unto Him, and you will be safe
from all evil.

Consider the Psalm on mishaps, and see how the author describes
that [390]great Providence, the protection and defence from all
mishaps that concern the body, both from those that are common to
all people, and those that concern only one certain individual; from
those that are due to the laws of Nature, and those that are caused
by our fellow-men. The Psalmist says: “Surely he will deliver thee
from the snare of the fowler, and from the noisome pestilence. He
shall cover thee with his feathers, and under his wings shalt thou
trust: His truth shall be thy shield and buckler. Thou shalt not be
afraid for the terror by night; nor for the arrow that flieth by day”
(Ps. xci. 3–5). The author then relates how God protects us from the
troubles caused by men, saying, If you happen to meet on your way
with an army fighting with drawn swords, killing thousands at your
left hand and myriads at your right hand, you will not suffer any
harm; you will behold and see how God judges and punishes the
wicked that are being slain, whilst you remain unhurt. “A thousand
shall fall at thy side, and ten thousand at thy right hand; but it shall
not come nigh thee. Only with thine eyes shalt thou behold and see
the reward of the wicked” (ibid. vers. 7, 8). The author then
continues his description of the divine defence and shelter, and
shows the cause of this great protection, saying that such a man is
well guarded “Because he hath set his love upon me, therefore will I
deliver him: I will set him on high, because he hath known my
name” (ibid. ver. 14). We have shown in previous chapters that by
the “knowledge of God’s name,” the knowledge of God is meant. The
above passage may therefore be paraphrased as follows: “This man
is well guarded, because he hath known me, and then (bi chashak)
loved me.” You know the difference between the two Hebrew terms
that signify “to love,” ahab and ḥashak. When a man’s love is so
intense that his thought is exclusively engaged with the object of his
love, it is expressed in Hebrew by the term ḥashak.

The philosophers have already explained how the bodily forces of

man in his youth prevent the development of moral principles. In a
greater measure this is the case as regards the purity of thought
which man attains through the perfection of those ideas that lead
him to an intense love of God. Man can by no means attain this so
long as his bodily humours are hot. The more the forces of his body
are weakened, and the fire of passion quenched, in the same
measure does man’s intellect increase in strength and light; his
knowledge becomes purer, and he is happy with his knowledge.
When this perfect man is stricken in age and is near death, his
knowledge mightily increases, his joy in that knowledge grows
greater, and his love for the object of his knowledge more intense,
and it is in this great delight that the soul separates from the body.
To this state our Sages referred, when in reference to the death of
Moses, Aaron, and Miriam, they said that death was in these three
cases nothing but a kiss. They say thus: We learn from the words,
“And Moses the servant of the Lord died there in the land of Moab
by the mouth of the Lord” (Deut. xxxiv. 5), that his death was a kiss.
The same expression is used of Aaron: “And Aaron the priest went
up into Mount Hor … by the mouth of the Lord, and died there”
(Num. xxxiii. 38). Our Sages said that the same was the case with
Miriam; but the phrase “by the mouth of the Lord” is not employed,
because it was not considered appropriate to use these words in the
description of her death as she was a female. The meaning of this
saying is that these three died in the midst of [391]the pleasure
derived from the knowledge of God and their great love for Him.
When our Sages figuratively call the knowledge of God united with
intense love for Him a kiss, they follow the well-known poetical
diction “Let him kiss me with the kisses of his mouth” (Song i. 2).
This kind of death, which in truth is deliverance from death, has
been ascribed by our Sages to none but to Moses, Aaron, and
Miriam. The other prophets and pious men are beneath that degree;
but their knowledge of God is strengthened when death approaches.
Of them Scripture says, “Thy righteousness shall go before thee; the
glory of the Lord shall be thy rereward” (Isa. lviii. 8). The intellect of
these men remains then constantly in the same condition, since the
obstacle is removed that at times has intervened between the
intellect and the object of its action; it continues for ever in that
great delight, which is not like bodily pleasure. We have explained
this in our work, and others have explained it before us.

Try to understand this chapter, endeavour with all your might to

spend more and more time in communion with God, or in the
attempt to approach Him; and to reduce the hours which you spend
in other occupations, and during which you are not striving to come
nearer unto Him. This instruction suffices for the object of this
treatise.

[Contents]
 CHAPTER LII
We do not sit, move, and occupy ourselves when we are alone and
at home in the same manner as we do in the presence of a great
king; we speak and open our mouth as we please when we are with
the people of our own household and with our relatives, but not so
when we are in a royal assembly. If we therefore desire to attain
human perfection, and to be truly men of God, we must awake from
our sleep, and bear in mind that the great king that is over us, and
is always joined to us, is greater than any earthly king, greater than
David and Solomon. The king that cleaves to us and embraces us is
the Intellect that influences us, and forms the link between us and
God. We perceive God by means of that light that He sends down
unto us, wherefore the Psalmist says, “In Thy light shall we see
light” (Ps. xxxvi. 9): so God looks down upon us through that same
light, and is always with us beholding and watching us on account of
this light. “Can any hide himself in secret places that I shall not see
him?” (Jer. xxiii. 24). Note this particularly.

When the perfect bear this in mind, they will be filled with fear of
God, humility, and piety, with true, not apparent, reverence and
respect of God, in such a manner that their conduct, even when
alone with their wives or in the bath, will be as modest as they are
in public intercourse with other people. Thus it is related of our
renowned Sages that even in their sexual intercourse with their
wives they behaved with great modesty. They also said, “Who is
modest? He whose conduct in the dark night is the same as in the
day.” You know also how much they warned us not to walk proudly,
since “the fulness of the whole earth is His glory” (Isa. vi. 3). They
thought that by these rules the above-mentioned idea will be firmly
established in the hearts of men, viz., that we are always before
God, and it is in the presence of His glory that we go to and fro. The
great men among our Sages would not uncover their heads because
they believed that God’s glory was round them [392]and over them;
for the same reason they spoke little. In our Commentary on the
Sayings of the Fathers (chap. i. 17) we have fully explained how we
have to restrict our speech. Comp. “For God is in heaven and thou
upon earth, therefore let thy words be few” (Eccles. v. 1).

What I have here pointed out to you is the object of all our religious
acts. For by [carrying out] all the details of the prescribed practices,
and repeating them continually, some few pious men may attain
human perfection. They will be filled with respect and reverence
towards God; and bearing in mind who is with them, they will
perform their duty. God declares in plain words that it is the object
of all religious acts to produce in man fear of God and obedience to
His word—the state of mind which we have demonstrated in this
chapter for those who desire to know the truth, as being our duty to
seek. Comp. “If thou wilt not observe to do all the words of this law
that are written in this book, that thou mayest fear this glorious and
fearful name, the Lord thy God” (Deut. xxviii. 58). Consider how
clearly it is stated here that the only object and aim of “all the words
of this law” is to [make man] fear “the glorious and fearful name.”
That this end is attained by certain acts we learn likewise from the
phrase employed in this verse: “If thou wilt not observe to do … that
thou mayest fear.” For this phrase clearly shows that fear of God is
inculcated [into our hearts] when we act in accordance with the
positive and the negative precepts. But the truths which the Law
teaches us—the knowledge of God’s Existence and Unity—create in
us love of God, as we have shown repeatedly. You know how
frequently the Law exhorts us to love God. Comp. “And thou shalt
love the Lord thy God with all thine heart, and with all thy soul, and
with all thy might” (Deut. vi. 5). The two objects, love and fear of
God, are acquired by two different means. The love is the result of
the truths taught in the Law, including the true knowledge of the
Existence of God; whilst fear of God is produced by the practices
prescribed in the Law. Note this explanation.

[Contents]