Introduction to Computer Graphics with OpenGL ES

First Edition Junghyun Han download

https://textbookfull.com/product/introduction-to-computer-
graphics-with-opengl-es-first-edition-junghyun-han/

Download more ebook from https://textbookfull.com

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

Computer Graphics Programming in OpenGL with Java

Gordon

https://textbookfull.com/product/computer-graphics-programming-
in-opengl-with-java-gordon/

Computer graphics through openGL Third Edition Guha

https://textbookfull.com/product/computer-graphics-through-
opengl-third-edition-guha/

Computer Graphics Through OpenGL: From Theory to

Experiments Sumanta Guha

https://textbookfull.com/product/computer-graphics-through-
opengl-from-theory-to-experiments-sumanta-guha/

Computer Graphics Programming in OpenGL with Java 2nd

Edition V. Scott Gordon

https://textbookfull.com/product/computer-graphics-programming-
in-opengl-with-java-2nd-edition-v-scott-gordon/
Computer Graphics Programming in OpenGL Using C++,
Third Edition Gordon Phd

https://textbookfull.com/product/computer-graphics-programming-
in-opengl-using-c-third-edition-gordon-phd/

Introduction to mechanism design: with computer

applications First Edition B. Dyer

https://textbookfull.com/product/introduction-to-mechanism-
design-with-computer-applications-first-edition-b-dyer/

Introduction to Computer Graphics A Practical Learning

Approach 1st Edition Fabio Ganovelli

https://textbookfull.com/product/introduction-to-computer-
graphics-a-practical-learning-approach-1st-edition-fabio-
ganovelli/

Learn OpenGL Learn modern OpenGL graphics programming

in a step by step fashion 1st Edition Joey De Vries

https://textbookfull.com/product/learn-opengl-learn-modern-
opengl-graphics-programming-in-a-step-by-step-fashion-1st-
edition-joey-de-vries/

Computer graphics with Open GL. 4th Edition Donald D.

Hearn

https://textbookfull.com/product/computer-graphics-with-open-
gl-4th-edition-donald-d-hearn/
 Introduction to
Computer Graphics
with OpenGL ES
 Introduction to
Computer Graphics
with OpenGL ES

JungHyun Han
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742

© 2018 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: 20180424

International Standard Book Number-13: 978-1-4987-4892-6 (Hardback)

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
Dedication

To my wonderful boy, Jihoon, who is embarking on a new journey in his life.

Contents

Preface xiii

Part I - Rendering Pipeline

1 Introduction 1
1.1 Computer Graphics Production . . . . . . . . . . . . . . . . 1
1.2 Graphics API . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 Mathematics: Basics 7
2.1 Matrices and Vectors . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Coordinate System and Basis . . . . . . . . . . . . . . . . . . 9
2.3 Dot Product . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.4 Cross Product . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.5 Line, Ray, and Linear Interpolation . . . . . . . . . . . . . . 13

3 Modeling 17
3.1 Polygon Mesh . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.1.1 Polygon Mesh Creation∗ . . . . . . . . . . . . . . . . . 19
3.1.2 Polygon Mesh Representation . . . . . . . . . . . . . . 23
3.2 Surface Normals . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2.1 Triangle Normals . . . . . . . . . . . . . . . . . . . . . 24
3.2.2 Vertex Normals . . . . . . . . . . . . . . . . . . . . . . 25
3.3 Polygon Mesh Export and Import . . . . . . . . . . . . . . . 26

4 Spaces and Transforms 31

4.1 2D Transforms and Matrix Representations . . . . . . . . . . 31
4.1.1 Scaling . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.1.2 Rotation . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.1.3 Translation and Homogeneous Coordinates . . . . . . 33
4.1.4 Composition of 2D Transforms . . . . . . . . . . . . . 35
4.2 Affine Transform . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.3 3D Transforms and Matrix Representations . . . . . . . . . . 39
4.3.1 Scaling . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.3.2 Rotation . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.3.3 Translation and Homogeneous Coordinates . . . . . . 41

vii
viii Contents

4.4 Application: World Transform . . . . . . . . . . . . . . . . . 42

4.5 Rotation and Object-space Basis . . . . . . . . . . . . . . . . 46
4.6 Inverse Transforms . . . . . . . . . . . . . . . . . . . . . . . . 48

5 Vertex Processing 53
5.1 World Transform Revisited . . . . . . . . . . . . . . . . . . . 54
5.2 View Transform . . . . . . . . . . . . . . . . . . . . . . . . . 57
5.2.1 Camera Space . . . . . . . . . . . . . . . . . . . . . . 57
5.2.2 View Matrix for Space Change . . . . . . . . . . . . . 58
5.3 Right-hand System versus Left-hand System . . . . . . . . . 60
5.4 Projection Transform . . . . . . . . . . . . . . . . . . . . . . 62
5.4.1 View Frustum . . . . . . . . . . . . . . . . . . . . . . . 62
5.4.2 Projection Matrix and Clip Space . . . . . . . . . . . 63
5.4.3 Derivation of Projection Matrix∗ . . . . . . . . . . . . 67

6 OpenGL ES and Shader 75

6.1 OpenGL ES and Shading Language . . . . . . . . . . . . . . 75
6.2 Vertex Shader . . . . . . . . . . . . . . . . . . . . . . . . . . 75
6.3 OpenGL ES for Shaders . . . . . . . . . . . . . . . . . . . . . 77
6.4 Attributes and Uniforms . . . . . . . . . . . . . . . . . . . . 78
6.4.1 Attributes and Buffer Objects . . . . . . . . . . . . . . 79
6.4.2 Uniforms . . . . . . . . . . . . . . . . . . . . . . . . . 82
6.5 Drawcalls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

7 Rasterizer 87
7.1 Clipping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
7.2 Perspective Division . . . . . . . . . . . . . . . . . . . . . . . 88
7.3 Back-face Culling . . . . . . . . . . . . . . . . . . . . . . . . 89
7.3.1 Concept . . . . . . . . . . . . . . . . . . . . . . . . . . 90
7.3.2 Implementation . . . . . . . . . . . . . . . . . . . . . . 90
7.4 Viewport Transform . . . . . . . . . . . . . . . . . . . . . . . 92
7.5 Scan Conversion . . . . . . . . . . . . . . . . . . . . . . . . . 94

8 Image Texturing 103

8.1 Texture Coordinates . . . . . . . . . . . . . . . . . . . . . . . 103
8.2 Surface Parameterization . . . . . . . . . . . . . . . . . . . . 106
8.3 Texture Definition in GL . . . . . . . . . . . . . . . . . . . . 108
8.4 Texture Wrapping . . . . . . . . . . . . . . . . . . . . . . . . 109
8.5 Texture Filtering . . . . . . . . . . . . . . . . . . . . . . . . . 111
8.5.1 Magnification . . . . . . . . . . . . . . . . . . . . . . . 113
8.5.2 Minification . . . . . . . . . . . . . . . . . . . . . . . . 113
8.6 Mipmapping . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
8.6.1 Mipmap Construction . . . . . . . . . . . . . . . . . . 114
8.6.2 Mipmap Filtering . . . . . . . . . . . . . . . . . . . . . 115
8.7 Texture Filtering in GL . . . . . . . . . . . . . . . . . . . . . 118
 Contents ix

8.8 Mipmappig Examples in GL∗ . . . . . . . . . . . . . . . . . . 118

8.9 Fragment Shader for Texturing . . . . . . . . . . . . . . . . . 121

9 Lighting 127
9.1 Phong Lighting Model . . . . . . . . . . . . . . . . . . . . . . 127
9.1.1 Diffuse Reflection . . . . . . . . . . . . . . . . . . . . . 128
9.1.2 Specular Reflection . . . . . . . . . . . . . . . . . . . . 130
9.1.3 Ambient Reflection . . . . . . . . . . . . . . . . . . . . 132
9.1.4 Emissive Light . . . . . . . . . . . . . . . . . . . . . . 132
9.2 Shaders for Phong Lighting . . . . . . . . . . . . . . . . . . . 133

10 Output Merger 139

10.1 Z-buffering . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
10.2 Alpha Blending . . . . . . . . . . . . . . . . . . . . . . . . . 142

Part II - Advanced Topics

11 Euler Transforms and Quaternions 149
11.1 Euler Transforms . . . . . . . . . . . . . . . . . . . . . . . . 149
11.1.1 World-space Euler Transforms . . . . . . . . . . . . . 149
11.1.2 Object-space Euler Transforms∗ . . . . . . . . . . . . 150
11.2 Euler Transforms for Keyframe Animation . . . . . . . . . . 153
11.2.1 2D Keyframe Animation . . . . . . . . . . . . . . . . . 153
11.2.2 3D Keyframe Animation . . . . . . . . . . . . . . . . . 154
11.2.3 Interpolation of Euler Angles . . . . . . . . . . . . . . 156
11.3 Quaternions . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
11.3.1 Quaternion Representation . . . . . . . . . . . . . . . 158
11.3.2 Rotations Using Quaternions . . . . . . . . . . . . . . 158
11.3.3 Interpolation of Quaternions . . . . . . . . . . . . . . 162
11.3.4 Conversion between Quaternion and Rotation Matrix 163

12 Screen-space Object Manipulation 167

12.1 Picking an Object . . . . . . . . . . . . . . . . . . . . . . . . 167
12.1.1 Screen-space Ray . . . . . . . . . . . . . . . . . . . . . 168
12.1.2 Camera-space Ray . . . . . . . . . . . . . . . . . . . . 168
12.1.3 Object-space Ray . . . . . . . . . . . . . . . . . . . . . 170
12.1.4 Intersection between Ray and Bounding Volume . . . 171
12.1.5 Intersection between Ray and Triangle . . . . . . . . . 175
12.2 Rotating an Object . . . . . . . . . . . . . . . . . . . . . . . 178

13 Character Animation 185

13.1 Skeleton and Space Change . . . . . . . . . . . . . . . . . . . 185
13.1.1 Skeleton . . . . . . . . . . . . . . . . . . . . . . . . . . 185
13.1.2 Space Change between Bones . . . . . . . . . . . . . . 186
x Contents

13.1.3 Character Space to Bone Space . . . . . . . . . . . . . 189

13.2 Forward Kinematics . . . . . . . . . . . . . . . . . . . . . . . 192
13.3 Skinning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
13.3.1 Vertex Blending . . . . . . . . . . . . . . . . . . . . . 196
13.3.2 Integration with Keyframe Animation∗ . . . . . . . . 198
13.4 Inverse Kinematics . . . . . . . . . . . . . . . . . . . . . . . . 200
13.4.1 Analytic Solution . . . . . . . . . . . . . . . . . . . . . 201
13.4.2 Cyclic Coordinate Descent . . . . . . . . . . . . . . . . 202

14 Normal Mapping 209

14.1 Height Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
14.2 Normal Map . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
14.3 Shaders for Normal Mapping . . . . . . . . . . . . . . . . . . 214
14.4 Tangent-space Normal Mapping . . . . . . . . . . . . . . . . 218
14.4.1 Tangent-space Normals . . . . . . . . . . . . . . . . . 218
14.4.2 Shaders for Tangent-space Normal Mapping . . . . . . 219
14.4.3 Computing Tangent Spaces∗ . . . . . . . . . . . . . . 222
14.5 Authoring Normal Maps . . . . . . . . . . . . . . . . . . . . 223

15 Shadow Mapping 231

15.1 Two-pass Algorithm . . . . . . . . . . . . . . . . . . . . . . . 231
15.2 Shadow Map Filtering . . . . . . . . . . . . . . . . . . . . . . 234
15.3 GL Program and Shaders for Shadow Mapping∗ . . . . . . . 236
15.3.1 First-pass Shaders . . . . . . . . . . . . . . . . . . . . 236
15.3.2 Render-to-Texture and Framebuffer Object . . . . . . 238
15.3.3 Second-pass Shaders . . . . . . . . . . . . . . . . . . . 239
15.4 Hard Shadow versus Soft Shadow . . . . . . . . . . . . . . . 242

16 Texturing toward Global Illumination 249

16.1 Global Illumination . . . . . . . . . . . . . . . . . . . . . . . 250
16.1.1 Ray Tracing . . . . . . . . . . . . . . . . . . . . . . . . 250
16.1.2 Radiosity∗ . . . . . . . . . . . . . . . . . . . . . . . . . 254
16.2 Light Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . 258
16.3 Environment Mapping . . . . . . . . . . . . . . . . . . . . . . 259
16.3.1 Cube Mapping . . . . . . . . . . . . . . . . . . . . . . 261
16.3.2 GL Program and Shaders for Cube Mapping . . . . . 262
16.4 Ambient Occlusion . . . . . . . . . . . . . . . . . . . . . . . 264

17 Parametric Curves and Surfaces 271

17.1 Parametric Curves . . . . . . . . . . . . . . . . . . . . . . . . 271
17.1.1 Bézier Curves . . . . . . . . . . . . . . . . . . . . . . . 271
17.1.2 Hermite Curve and Catmull-Rom Spline . . . . . . . . 275
17.1.3 Application: Camera Path . . . . . . . . . . . . . . . . 277
17.2 Parametric Surfaces . . . . . . . . . . . . . . . . . . . . . . . 279
17.2.1 Bilinear Patch . . . . . . . . . . . . . . . . . . . . . . 280
 Contents xi

17.2.2 Biquadratic Bézier Patch . . . . . . . . . . . . . . . . 282

17.2.3 Bicubic Bézier Patch . . . . . . . . . . . . . . . . . . . 286
17.2.4 Bézier Triangle . . . . . . . . . . . . . . . . . . . . . . 288

18 Surface Tessellation 295

18.1 Displacement Mapping . . . . . . . . . . . . . . . . . . . . . 295
18.1.1 GPU Tessellation . . . . . . . . . . . . . . . . . . . . . 295
18.1.2 Shaders and Tessellator . . . . . . . . . . . . . . . . . 297
18.2 PN-triangles∗ . . . . . . . . . . . . . . . . . . . . . . . . . . 303
18.2.1 Computing Control Points . . . . . . . . . . . . . . . . 303
18.2.2 Computing Control Normals . . . . . . . . . . . . . . 308
18.2.3 GPU Tessellation . . . . . . . . . . . . . . . . . . . . . 310
18.2.4 Shaders and Tessellator . . . . . . . . . . . . . . . . . 311

References 319

Index 321
Preface

OpenGL ES is the standard graphics API for mobile and embedded sys-
tems. Virtually every pixel on a smartphone’s screen is generated by OpenGL
ES. However, there exists no textbook on OpenGL ES which has a balance be-
tween theory and practicality. This book is written to answer that need and
presents the must-know in real-time graphics with OpenGL ES. This book
suits the advanced undergraduate and beginner graduate courses in computer
graphics.
Another primary group of readers that this book may benefit includes mo-
bile 3D app developers, who have experiences in OpenGL ES and shader
programming but lack theoretical background in 3D graphics. A few excel-
lent programming manuals on OpenGL ES can be found in bookstores, but
they do not provide a sufficient level of mathematical background for devel-
opers. Assuming that the readers have a minimal understanding of vectors
and matrices, this book provides an opportunity to combine their knowledge
with the background theory of computer graphics.
This book is built upon the author’s previous work 3D Graphics for Game
Programming published in 2011. Reusing roughly half of the contents from
that book, several new topics and a considerable number of OpenGL ES and
shader programs have been added. As OpenGL ES is a subset of OpenGL,
this book is also suitable for beginner OpenGL programmers.
The organization and presentation of this book have been carefully designed
so as to enable the readers to easily understand the key aspects of real-time
graphics and OpenGL ES. Over the chapters, numerous 3D illustrations are
provided to help the readers effortlessly grasp the complicated topics. An
important organizational feature of this book is that “non-core” details are
presented in separate notes (in shaded boxes) and in optional sections (marked
by asterisks). They can be safely skipped without incurring any difficulty in
understanding the subsequent topics of the book.
If the optional parts are excluded, the entire contents of this book can be
covered in a 16-week semester for graduate classes. For undergraduate classes,
however, this feat will be difficult. According to the author’s experience,
teaching Chapters 1 through 14 is a feasible goal.
The sample programs presented in this book are available on GitHub:
https://github.com/medialab-ku/openGLESbook. The site also provides links
to the full-length lecture notes as PowerPoint files and additional materials
including video clips.

xiii
xiv Preface

Acknowledgments

The author is deeply indebted to his students at the 3D Interactive Media

Lab of Korea University. Virtually all visual illustrations were generated by
Myoung Gon Kim and Seungjik Lee, who have exceptional talents in both
programming and visual arts. Inbum Park performed a lot of 3ds Max works
requested by the author, and JeongHyeon Ahn was in charge of all sample
programs. HyeongYeop Kang consistently helped the author reorganize the
chapters, and Paul C. Gloumeau proofread the beta version of this book.
The other lab members, Seungho Baek, SungIk Cho, Sang-bin Kim, Seung-
wook Kim, Jinwoo Choi, Geonsun Lee, Min Hyung Kee, Seongsu Kwon, and
Sun Young Park, always took time out of their busy schedules to respond
to my inquiries in real time. The content of this book has been gradually
built through the courses offered at Korea University. The students of the
classes provided the author with invaluable feedback with which the book
was improved.
My greatest appreciation goes to the brightest faces I have met in my life,
Kyung-Ok, Jeehee, and Jihoon. Thank you for always being with me. I love
you so much.

JungHyun Han
Computer Science Department
Korea University
Seoul, Korea
 Part I
Rendering Pipeline
Chapter 1
Introduction

Computer graphics refers to the process of generating images using com-

puters. Three-dimensional (3D) computer graphics takes as input 3D repre-
sentations of objects and performs various calculations on them to produce
images. The last three decades have seen a profusion of 3D computer graphics
in films and video games.
Computer-generated images are often called frames. An illusion of move-
ment is generated on the screen by displaying a sequence of changing frames.
The frames can be generated at real time. The best-known area of real-time
graphics is video games, which typically produce more than 30 frames per
second (fps). On the other hand, special effects in films often take as long as
minutes or hours for a single frame. In return, we may obtain images that are
hardly distinguishable from views of the real world. The algorithms and tech-
niques adopted in real-time graphics are fairly different from those in off-line
graphics. This book presents the essential components of real-time graphics.

1.1 Computer Graphics Production

Computer graphics production is often described in five steps shown in
Fig. 1.1, where graphic artists and programmers are the key players. Model-
ing, rigging, and animation are off-line tasks performed by the artists. At run
time, computer programs replay the animation and perform rendering and
post-processing.
A model is referred to as a computer representation of an object, and model-
ing is the process of creating the objects comprising the virtual scenes. Con-

Fig. 1.1: The major steps in computer graphics production.

1
2 Introduction to Computer Graphics with OpenGL ES

Fig. 1.2: Almost all 3D models in real-time graphics are represented in polygon
meshes.

sider a baseball game. We need players, bats, balls, etc. They are usually
represented in polygons, as shown in Fig. 1.2. Such polyhedral objects are
named the polygon meshes.
The scope of modeling is not limited to constructing 3D models but includes
creating textures. The simplest form of a texture is an image that is pasted
on an object’s surface. Fig. 1.3-(a) shows an image texture created for the
baseball player model. The texture is pasted on the surface of the player at
run time to produce the result shown in Fig. 1.3-(b).
The baseball player should be able to hit a ball, run, and slide into a base,
i.e., we need to animate the player. For this purpose, we usually specify
the skeleton or rig of the player. Fig. 1.4 shows a skeleton embedded in the
polygon model. We then define how the skeletal motion deforms the player’s
polygon mesh such that, for example, the polygons of the arm are made to
move when the arm bone is lifted. This process is often referred to as rigging.
The graphics artist creates a sequence of skeletal motions. At run time, the
skeletal motions are replayed “per frame” and the polygon mesh is animated
over frames. Fig. 1.5 shows a few snapshots of an animated player.
Rendering is the process of generating a 2D image from a 3D scene. The
image makes up a frame. Fig. 1.6 shows the results of rendering the dynamic
scene of Fig. 1.5. Realistic rendering is a complicated process, in which lighting
as well as texturing is an essential component. For example, the shadow shown
in Fig. 1.6 is a result of lighting.
The final step in the production of computer graphics, post-processing, is
optional. It uses a set of special operations to give additional effects to the
rendered images. An example is motion blur shown in Fig. 1.7. When a
camera captures a scene, the resulting image represents the scene over a short
period of time. Consequently, rapidly moving objects may result in motion
 Introduction 3

Fig. 1.3: Image texturing example: (a) This texture is a collection of small
images, each of which is for a part of the baseball player’s body. The texture
may look weird at first glance, but Chapter 8 will present how such a texture
is created and used. (b) The texture is pasted on the player’s polygon mesh
at run time.

Fig. 1.4: A skeleton is composed of bones and is embedded in the polygon mesh
for animation. This figure illustrates the bones as if they were solids, but the
bones do not have explicit geometric representations. They are conceptual
entities that are usually represented as matrices. This will be detailed in
Chapter 13.

blur. It is a desirable effect in computer graphics, without which the moving

objects would appear staggered or jerky. Another example is the depth-of-field
effect, which makes objects outside of the camera’s focus appear blurry.
4 Introduction to Computer Graphics with OpenGL ES

Fig. 1.5: The polygon mesh can be animated by controlling the skeleton
embedded in it.

Fig. 1.6: Results of rendering the animated player.

Fig. 1.7: Motion blur effects.

 Introduction 5

1.2 Graphics API

The artists perform modeling, rigging, and animation in an off-line mode.
Dedicated programs such as Autodesk 3ds Max and Autodesk Maya are pop-
ularly used. In contrast, run-time animation, rendering, and post-processing
are executed by an application program. In games, the application is typi-
cally built upon a game engine. See Fig. 1.8. A game engine is a development
tool that provides a suite of indispensable modules for animation, rendering,
and post-processing. Modern game engines provide additional modules for
physics-based simulation, sound, artificial intelligence, and so on. Well-known
game engines include Unity Technologies’ Unity and Epic Games’ Unreal En-
gine.
In general, a game engine is built upon graphics APIs (Application Pro-
gramming Interfaces). Two popular APIs are Direct3D [1] and OpenGL
(Open Graphics Library) [2]. Direct3D is part of Microsoft’s DirectX API
and is available only for Microsoft platforms. OpenGL is a cross-platform
API managed by a non-profit consortium for open standards, the Khronos
Group. OpenGL ES (OpenGL for Embedded Systems) is a 3D API for hand-
held and embedded devices and is defined as a subset of OpenGL. It is widely
used for mobile graphics programming.

Fig. 1.8: Software and hardware hierarchy for mobile games.

Graphics APIs provide 3D applications or game engines with essential

graphics functions. Today, such functions are implemented in GPU (Graphics
Processing Unit). It is a processor specialized for graphics. A graphics API
can be taken as the software interface of the GPU. The API translates the
application’s graphics commands to instructions that can be executed by the
GPU.
Chapter 2
Mathematics: Basics

This chapter delivers an explicit presentation of the basic mathematical

techniques, which are needed throughout this book. Matrices and vectors are
indispensable tools in computer graphics and we start with them.

2.1 Matrices and Vectors

Shown below is a matrix with m rows and n columns:
 
a11 a12 · · · a1n
 a21 a22 · · · a2n 
(2.1)
 
 .. .. . . .. 
 . . . . 
am1 am2 · · · amn
Its dimensions are denoted as m×n. If m = n, the matrix is called square.
Two subscripts of an element specify where it is located. For instance, a12 is in
the first row and second column. Two matrices A and B can be multiplied if
the number of columns in A equals the number of rows in B. If A’s dimensions
are l×m and B’s dimensions are m×n, AB is an l×n matrix. See the example
shown below:
 
a11 a12  
b b b
AB =  a21 a22  11 12 13
b21 b22 b23
 a31 a32  (2.2)
a11 b11 + a12 b21 a11 b12 + a12 b22 a11 b13 + a12 b23
=  a21 b11 + a22 b21 a21 b12 + a22 b22 a21 b13 + a22 b23 
a31 b11 + a32 b21 a31 b12 + a32 b22 a31 b13 + a32 b23
A 2D vector is usually represented as (x, y) and a 3D vector as (x, y, z).
They are often called the row vectors. Alternatively, we can use the column
vector representation, e.g., a 2D column vector is written as
 
x
(2.3)
y
Vectors are special matrices. The above vector is a matrix with a single
column and its dimensions are 2×1. Then, the method of matrix-matrix

7
8 Introduction to Computer Graphics with OpenGL ES

multiplication applies to matrix-vector multiplication. If M is a 3×2 matrix

and v is a 2D column vector, for example, M v is computed as follows:
 
ab  
x
Mv =  c d 
y
e f  (2.4)
ax + by
=  cx + dy 
ex + f y

Given a matrix M , its transpose denoted by M T is obtained by interchang-

ing the rows and columns of M . For example, the transpose of M given in
Equation (2.4) is defined as  
ac e
(2.5)
bdf
The same applies to vectors. If v is a column vector, its transpose, v T , is the
row-vector counterpart.
The matrix-vector multiplication in Equation (2.4) can be represented in a
different way. Instead of the column vector, v, we can use the row vector, v T ,
and place it at the left of M T :
 
T T

ac e
v M = xy
bdf
(2.6)
= ax + by cx + dy ex + f y

The result is the same as in Equation (2.4) but is represented in a row vector.
(Whereas OpenGL uses the column vectors and the vector-on-the-right repre-
sentation for matrix-vector multiplication, Direct3D uses the row vectors and
the vector-on-the-left representation.)
The identity matrix is a square matrix with ones on the main diagonal (from
the upper-left element to the lower-right element) and zeros everywhere else.
It is denoted by I. For any matrix M , M I = IM = M , as shown in the
following examples:
    
a11 a12 a13 100 a11 a12 a13
 a21 a22 a23   0 1 0  =  a21 a22 a23  (2.7)
a31 a32 a33 001 a31 a32 a33
    
100 a11 a12 a13 a11 a12 a13
 0 1 0   a21 a22 a23  =  a21 a22 a23  (2.8)
001 a31 a32 a33 a31 a32 a33
If two square matrices A and B are multiplied to return an identity matrix,
i.e., if AB = I, B is called the inverse of A and is denoted by A−1 . By
the same token, A is the inverse of B. Note that (AB)−1 = B −1 A−1 as
(AB)(B −1 A−1 ) = A(BB −1 )A−1 = AIA−1 = AA−1 = I. Similarly, (AB)T =
B T AT .
 Mathematics: Basics 9

The coordinates of a 2D vector q

v are represented by (vx , vy ). Its length
2 2
denoted by kvk is defined q
as v x + v y . If v is a 3D vector, its coordinates are
(vx , vy , vz ) and kvk is vx2 + vy2 + vz2 . Dividing a vector by its length is called
normalization. The resulting vector, v/kvk, has the same direction as v. Such
a normalized vector is called the unit vector since its length is one.

2.2 Coordinate System and Basis

Fig. 2.1: Basis examples: (a) Standard basis. (b) A valid basis that is neither
standard nor orthonormal. (c) An orthonormal basis that is not standard.

In the 2D coordinate system shown in Fig. 2.1-(a), e1 = (1, 0) along the x -

axis and e2 = (0, 1) along the y-axis. (Throughout this book, we use the terms
coordinate system and space interchangeably.) In the 2D space, every vector
can be defined as a linear combination 1 of e1 and e2 , e.g., (3, 5) = 3e1 + 5e2 .
In this sense, {e1 , e2 } is a basis. Specifically, it is named the standard basis.
Since e1 and e2 are unit vectors that are orthogonal to each other, they are
also called orthonormal .
A different set of vectors may work as a basis. Consider {(1, 1), (0, 2)} shown
in Fig. 2.1-(b). Our example vector, (3, 5), can be defined by linearly combin-
ing (1, 1) and (0, 2), i.e., (3, 5) = 3(1, 1) + 1(0, 2). Observe that {(1, 1), (0, 2)}
is not an orthonormal basis. It is generally easier to work with an orthonormal
basis rather than an arbitrary basis. Fig. 2.1-(c)
√ shows
√ an orthonormal basis,
which is not standard. Note that (3, 5) = 4 2u + 2v.

1 Given n vectors, v1 , v2 , . . . , vn , and n scalars, c1 , c2 , . . . , cn , the linear combination of the

vectors with the scalars is c1 v1 + c2 v2 + · · · + cn vn . It is also a vector.
10 Introduction to Computer Graphics with OpenGL ES

Fig. 2.2: Standard basis for the 3D coordinate system.

Fig. 2.3: Dot product of two vectors reveals their relative orientation.

As shown in Fig. 2.2, the standard basis in the 3D space is {e1 , e2 , e3 },

where e1 = (1, 0, 0), e2 = (0, 1, 0), and e3 = (0, 0, 1). It is an orthonormal
basis. All 2D and 3D bases presented from now on will be orthonormal.

2.3 Dot Product

Consider two n-dimensional vectors, a and b. When their coordinates are
represented by (a1 , a2 , . . . , an ) and (b1 , b2 , . . . , bn ), respectively, their dot prod-
uct, denoted by a · b, is defined as follows:
n
X
a·b= ai bi = a1 b1 + a2 b2 + . . . + an bn (2.9)
i=1

When the angle between a and b is denoted as θ, a · b can be also defined

as follows:
a · b = kakkbkcosθ (2.10)
 Mathematics: Basics 11

Fig. 2.4: Cross product and right-hand rule: (a) a × b. (b) b × a.

If a and b are orthogonal to each other, θ = 90◦ and a · b = 0, as shown on

the left of Fig. 2.3. See the other examples in Fig. 2.3. If θ is an acute angle,
a · b is positive; if θ is an obtuse angle, a · b is negative. The same observation
can be made when a and b are 3D vectors.
Equation (2.10) asserts that, if v is a unit vector, v · v = 1. An orthonormal
basis has an interesting and useful feature. Consider 2D and 3D standard
bases. Given the 2D standard basis, {e1 , e2 }, e1 · e1 = e2 · e2 = 1 and e1 · e2 =
e2 · e1 = 0. In the 3D standard basis, {e1 , e2 , e3 }, a similar observation is
made, i.e., ei · ej = 1 if i = j, and ei · ej = 0 otherwise. This feature applies
to every orthonormal basis.

2.4 Cross Product

The cross product takes two 3D vectors, a and b, and returns another 3D
vector which is perpendicular to both a and b, as shown in Fig. 2.4-(a). The
cross product is denoted by a × b and its direction is defined by the right-hand
rule. The direction of a × b is indicated by the thumb of the right hand when
the other fingers curl from a to b. The length of a × b equals the area of the
parallelogram that a and b span:

ka × bk = kakkbk sin θ (2.11)

The right-hand rule implies that the direction of b × a is opposite to that

of a × b, i.e., b × a = −(a × b). However, their lengths are the same, as
illustrated in Fig. 2.4-(b). In this sense, the cross product operation is called
anti-commutative. When a = (ax , ay , az ) and b = (bx , by , bz ), a × b = (ay bz −
az by , az bx − ax bz , ax by − ay bx ). See [Note: Derivation of cross product]2 .

2 You can skip the notes in shaded boxes. No trouble will be encountered in further reading.
12 Introduction to Computer Graphics with OpenGL ES

[Note: Derivation of cross product]

Fig. 2.5: Standard basis and right-hand rule: The thumb of the right hand
points toward e3 when the other four fingers curl from e1 to e2 , i.e., e1 × e2 =
e3 . Similarly, e2 × e3 = e1 and e3 × e1 = e2 .
In Fig. 2.5, the relative orientations among the basis vectors, e1 , e2 , and
e3 , are described using the right-hand rule:
e1 × e2 = e3
e2 × e3 = e1 (2.12)
e3 × e1 = e2
The anti-commutativity of the cross product leads to the following:
e2 × e1 = −e3
e3 × e2 = −e1 (2.13)
e1 × e3 = −e2
Equation (2.11) also asserts that
e1 × e1 = e2 × e2 = e3 × e3 = 0 (2.14)
where 0 is the zero vector, (0, 0, 0).
When a = (ax , ay , az ) and b = (bx , by , bz ), a is rewritten in terms of the
standard basis as ax e1 + ay e2 + az e3 . Similarly, b is rewritten as bx e1 + by e2 +
bz e3 . Then, a × b is derived as follows:
a × b = (ax e1 + ay e2 + az e3 ) × (bx e1 + by e2 + bz e3 )
= ax bx (e1 × e1 ) + ax by (e1 × e2 ) + ax bz (e1 × e3 )+
ay bx (e2 × e1 ) + ay by (e2 × e2 ) + ay bz (e2 × e3 )+
az bx (e3 × e1 ) + az by (e3 × e2 ) + az bz (e3 × e3 )
(2.15)
= ax bx 0 + ax by e3 − ax bz e2
−ay bx e3 + ay by 0 + ay bz e1
+az bx e2 − az by e1 + az bz 0
= (ay bz − az by )e1 + (az bx − ax bz )e2 + (ax by − ay bx )e3
The coordinates of a × b are (ay bz − az by , az bx − ax bz , ax by − ay bx ).
Other documents randomly have
different content
the brave Swiss Bouquet led the first English army that crossed the
Ohio river, making a tri-track road to the Muskingum valley and
bringing to a triumphal close Pontiac’s bloody rebellion. The old
Iroquois trail up the Mohawk valley and across the great watershed
of New York to the Niagara river was a famous Revolutionary
highway and afterward became one of the important pioneer routes.
On the Great Trail to Detroit Lachlan McIntosh erected the first fort
built by the thirteen colonies west of the Ohio, Fort Laurens on the
Muskingum near Great Crossings, where Bouquet had thrown his
army across the river in 1764. Indeed, throughout that whole half-
century of conflict in the Central West the lines of conquest were the
lines of the earlier routes of travel. Washington, Braddock, Forbes,
Bouquet, Lewis, Shirley, Sullivan, Clark, Brodhead, Crawford, Irvine,
McIntosh, Harmar, St. Clair, Wayne, and Harrison followed these old
highways and fought their battles on and beside them. These
campaigns were not made by water but by land. Had they been
made by rivers, the courses of their routes would have been
frequently described and mapped as having an important bearing on
the history of each campaign. Because they were made by land over
routes which have never received attention from historians the real
ground-work of these campaigns has been entirely omitted. Each
would be far better understood in every way if its route were clearly
defined. A thorough understanding of our history is impossible
without a knowledge of these highways of trade and war and the
strategic points they covered and connected.
But of vaster interest is the study of the surging armies of
pioneers and the occupation of the great empire conquered by these
armies for them. To the emigrant each tawny trail was a path to a
Promised Land. They came in thousands and hundreds of thousands
over the roads of New York, Pennsylvania, and Virginia. And what
roads they were! It was impossible for those pioneer wagons to
follow the Indian paths with any exactness. Even Braddock avoided
the steeper hills and yet was compelled to lower his wagons from
some hills with blocks and tackling—many being demolished at that.
And yet to avoid the high ground was inevitably to run into bogs and
swamps which were even worse than the hills. We do not have
roads a mile wide nowadays, but this was not an unheard-of thing in
the days of the pioneer roads. It was preferable to have them a mile
wide rather than a mile deep, which would certainly have been the
case in some places if one track had been used alone. And even with
numberless side tracks, skirting in every direction around the more
dangerous localities, horses were not infrequently drowned, and
great wagons heavy with freight sometimes sank completely out of
sight. The Black Swamp Road through Ohio south of Lake Erie was
one of the most important in the West. It is recorded that on one
occasion six horses were able to draw a two-wheeled vehicle but
fifteen miles in three days. A newspaper of August 31, 1837, affirms
that “the road through the Black Swamp has been much of the
season impassable. A couple of horses were lost in a mud hole last
week. The bottom had fallen out. The driver was unaware of the
fact. His horses plunged in and ere they could be extricated were
drowned.” It is comforting to think there has been some
improvement in our country highways. Such accounts as this would
have a tendency to influence the most skeptical.
The rivers were also great highways for emigration, particularly
such streams as the Ohio which flowed west. With the building of
the great canals new and more stable methods of travel were at the
disposal of prospective travelers and there was an increase in the
great tide of home-seekers. The smaller inland rivers were not likely
so largely used by these armies of pioneers as some have thought.
For instance, in a history of one of the interior counties of Ohio
(which is divided by one of the best rivers in the West) is a twenty-
five page description of the first immigrants, and of only one does it
say: “James Oglesby was a very early settler ... and is said to have
traveled up the Muskingum and Walhounding rivers in true Indian
style in a canoe.” And, though the Ohio river was always a great
highway to the West and Southwest, it was used less perhaps in the
early days of the immigration than later. Flat and keel boats cost
money, and money was a scarce article. In summer the river was
very low, and one party of pioneers, at least, spent one-third of the
entire time of journey from Connecticut to Marietta, Ohio, in coming
down the Ohio from near Pittsburg. It took half as long to come
those two hundred miles by river as to come all the way from
Connecticut to the Ohio in a cart drawn by oxen. Moreover, even as
late as the time of the starting of a regular line of steamer packets
from Pittsburg to Cincinnati (1796) the passengers were assured in
an advertisement that, in addition to being provided a place to sleep
and something to eat, they would have each a loophole from which
to shoot! The coming of steam navigation revolutionized river travel
as later it revolutionized land travel.

This series of monographs will treat of the history of America as

portrayed in the evolution of its highways of war and immigration
and commerce. The more important highways, both land and water,
will be specifically treated with reference to the national needs which
they temporarily or permanently satisfied. The study of any highway
for itself alone might prove of indifferent value, even though it were
an Appian Way; but the story of a road, which shows clearly the rise,
nature, and passing of a nation’s need for it, is of importance. It is
not of national import that there was a Wilderness Road to Kentucky,
but it is of utmost importance that a road through Cumberland Gap
made possible the early settlement of Kentucky, in that Kentucky
held the Mississippi river for the feeble colonies through days when
everything in the West and the whole future of the American
republic lay in a trembling balance. It is not of great importance that
there was a Nemacolin’s Path across the Alleghanies; but if for a
moment we can see the rough trail as the young Major Washington
saw it while the vanguard of the ill-fated Fry’s army marched out of
Wills Creek toward the Ohio, or if we can picture that terrible night’s
march Washington made from Fort Necessity when Jumonville’s
scouting party was run at last to cover by Half King’s Indians, we
shall know far better than ever the true story of the first campaign
of the war which won America for England, and realize as never
before what a brave, daring youth he was who on Indian trails
learned lessons that fitted him to become a leader of half-clothed, ill-
equipped armies.

The first aim of these monographs is suggestiveness; there is a

vast deal of geographic-historical work to be done throughout the
United States. There is no more interesting outdoor work for local
students than to trace, each in his own locality, the old land and
water highways, Indian trails, portage paths, pioneer roads or early
county or state roads. Maps should be made showing not only the
evolution of road-making in each county in the entire land, but all
springs and licks of importance should be correctly located and
mapped; sites of Indian villages should be marked; frontier forts and
blockhouses should be platted, including the surrounding defenses,
covered ways, springs or wells, and paths to and fro; traders’ huts
should all be placed, ancient boundary lines marked, old hunting-
grounds mapped. Those who can assist students in such
explorations are fast passing away. Much can be done this year that
can never be done so well in all the years which will succeed.
In subsequent monographs the author will endeavor to thank such
persons as have assisted him in their preparation. For the work
already completed and for that yet to be done I am especially
indebted to Mr. Arthur H. Clark, for encouragement and assistance;
to the Hon. Rodney Metcalf Stimson, for the freedom of his splendid
collection of Americana lately presented to Marietta College; and to
the patient, devoted assistance and collaboration of my wife.
A. B. H.
Marietta, Ohio, July 1, 1902.
PART I
Paths of Mound-Building Indians
 CHAPTER I
THE COMPARATIVE METHOD OF STUDY
The latest explorations of the mounds erected by those first
Americans, known best as the mound-building Indians, have
revolutionized our conceptions of the earliest race of which we know
in America. Very many notions, founded upon the authority of the
earlier archæologists, seem now to be either partly or wholly
incorrect. Many assumptions as to the population of this country
during the mound-building era, the degree of the civilization, and the
perfection of the arts, have not been substantiated by the more
accurate studies which have been made in the past decade.
The most important reason why so little progress has been made
in unraveling the mystery of the mounds that abound in central
North America is that “the authors who have written upon the
subject of American archæology have proceeded upon certain
assumptions which virtually closed the door against a free and
unbiased investigation.”[1] Of these assumptions, the one most
detrimental to the advance of the study of archæology is that which
has affirmed that “mound-builders” and American Indians were two
distinct races; thus all conclusions reached concerning the “mound-
builders” which were founded upon the earliest knowledge we have
of the American Indian were denied to archæology. But the evidence
of latest explorations and investigations makes it positive that the
“mound-builders” were not a race distinct and apart from the race
we know as American Indians: “The links directly connecting the
Indians and mound-builders are so numerous and well established
that archæologists are justified in accepting the theory that they are
one and the same people.”[2]
 This fact having been placed beyond the realm of speculation, a
great mass of testimony furnished by the study of the American
Indians is to be admitted in settlement of the question raised
concerning the history of the mound-builders in America.
First, this testimony will be found to set aside once and for all the
exaggerated statements as to the high grade of civilization reached
by these first Americans. It does not appear that the mound-building
Indians occupied a higher plane than that reached by the Indians as
first known by Europeans. One of the most exaggerated notions of
these Indians is that which ascribes to them very perfect ideas of
measurement; the alleged mathematical accuracy of certain of their
monumental works having been cited repeatedly as a sign of their
advanced stage of civilization. It has even been affirmed that their
mathematical accuracy could hardly be excelled by the most skillful
engineers of our day.
Recent explorations have dispelled this entertaining theory: “The
statement so often made that many of these monuments have been
constructed with such mathematical accuracy as to indicate not only
a unit of measure, but also the use of instruments, is found upon a
reëxamination to be without any basis, unless the near approach of
some three or four circles and as many squares of Ohio to
mathematical correctness be sufficient to warrant this opinion. As a
very general and in fact almost universal rule the figures are more or
less irregular, and indicate nothing higher in art than an Indian could
form with his eye and by pacing.”[3]
The fanciful theory of a great teeming population during the
mound-building era is equally without foundation. Even the size and
extent of the mounds do not imply a great population. An authority
of reputation gives figures from which it can be shown that four
thousand men, each transporting an equivalent of one wagon-load
of earth and stone a day, could have erected all the mounds in the
state of Ohio (which contains a greater number than any other in
the Union) in one generation.[4]
 When it is realized that the art for which the earliest Indians have
been most extolled was not, in reality, in advance of that known by
the ordinary Indian, and that the population of the country in the
mound-building era cannot be shown to have exceeded the
population found when the first white men visited the Indian races,
it is easy to see in the erection of the mounds, the burial of dead,
the rude implements common to both, the poor trinkets used for
ornamentation, the houses each built, the weapons each employed,
a vast deal of additional testimony proving that the “Mound-builder”
and Indian were of one race.
Thus the true historical method must be to compare what is
definitely known of the earliest Indians with the relics and memorials
which are surely those of the mound-building era in order to reach
undoubted facts concerning the prehistoric Indians. This applies
equally to customs, weapons, edifices, ornaments, and what is of
present moment to our study: highways of travel.
However, a complete detailed study of the highways of the early
Indians according to this method will not be indulged in for certain
respectable reasons; there are very few undoubted routes of the
mound-building Indians, and a detailed comparative study of these
and later Indian routes would become, under the circumstances, too
speculative to be of genuine historical value. Our purpose, then, will
be, merely to give the reasons for believing that the earliest Indians
had great overland routes of travel; that, though they lived largely in
the river valleys, their migrations were by land and not by water—in
short, that these first Americans undoubtedly marked out the first
highways of America used by man, as the large game animal, the
buffalo, marked out the first great highways used by animal life.
These highways were the highestways because their general
alignment was on the greater watersheds: and our study may be
better described perhaps by a subtitle—a study in highestways.
 CHAPTER II
DISTRIBUTION OF MOUND-BUILDING
INDIANS
The mounds of these first Americans of which we know are found
between Oregon and the Wyoming valley, in Pennsylvania, and
Onondaga county in New York; they extend from Manitoba in
Canada to the Gulf of Mexico.
The great seat of empire was in the drainage area of the
Mississippi river; on this river and its tributaries were the heaviest
mound-building populations. Few mounds are found east of the
Alleghany mountains.
In the Catalogue of Prehistoric Works East of the Rocky
Mountains, issued by the Bureau of Ethnology of the Smithsonian
Institution,[5] the geographical extent and density of the mounds in
central North America is brought out state by state with striking
suggestiveness. While the layman is warned that these maps
“present some features which are calculated to mislead,” and that
the maps indicate “to some extent the more thoroughly explored
areas rather than the true proportion of the ancient works in the
different sections,” still some conclusions have already been reached
which future exploitation will never weaken.
It is not expected that future investigation will change the verdict
that the heaviest mound-building population found its seat near the
Mississippi and Ohio rivers. “There is little doubt,” writes Dr. Thomas,
“that when Mississippi, Alabama, and Georgia have been thoroughly
explored many localities will be added to those indicated ... but it is
not likely that the number will be found to equal those in the area
drained by the Ohio and its affluents or in the immediate valley of
the Mississippi.”[6]
This fact, that the heaviest populations of the mound-building
Indians seem to have been near the Mississippi and Ohio is, of
course, shown by the archæological maps. In a rough way, subject
to the limitations previously mentioned, it can be found that the
following fourteen states contain evidences of having held the
heaviest ancient populations:

Ohio, Michigan,
Wisconsin, Georgia and
Tennessee, Arkansas,
Illinois, Missouri and
Florida, North
New York, Carolina,
Kentucky, Minnesota,
Indiana, Iowa,
Pennsylvania.

Now, by our last census the states which contain the largest
population today are:

New York, Indiana,

Pennsylvania, Michigan,
Illinois, Iowa,
Ohio, Georgia,
Texas, Kentucky,
Missouri, Wisconsin,
Massachusetts, Tennessee.

Thus of the fourteen most thickly populated states today perhaps

twelve give fair evidence of having been most thickly populated in
prehistoric times. As a general rule (but one growing less reliable
every day) the heaviest population has always been found in the
best agricultural regions; the states having the largest number of
fertile acres have had, as a rule, the largest populations—or did have
until the cities grew as they have in the past generation.
This argument, though necessarily loose, still is of interest and of
some importance in the present study. The earliest Indians found,
without any question, the best parts of the country they once
inhabited if we can take the verdict of the present race which
occupies the land.

Click here for larger image size

Archaeological Map of Wisconsin

[Showing interior location of remains]
 Coming down to a smaller scope of territory, can it be shown that
in the case of any one state the early Indians occupied the portions
most heavily populated today? It has been said that, in Ohio, four
counties contain evidence of having been the scenes of special
activity on the part of the earliest inhabitants: Butler, Licking, Ross,
and Franklin. These are interior counties (at a distance from the
Ohio and Lake Erie) and, of the remaining sixty-three interior
counties in the state, only seven exceeded these four in population
in 1880—when the cities had not so largely robbed the country
districts of their population as now. Thus the aborigines seem to
have been busiest where we have been busiest in the last half of the
nineteenth century.
In Wisconsin the mound-building Indians labored most in the
southern part of the state, where the bulk of that state’s population
is today—seventy-five per cent being found in the southern (and
smaller) half of the state.
In Michigan, a line drawn from the northern coast of Green Bay to
the southwestern corner of the state includes a very great
proportion of the archæological remains in the state. That line today
embraces on the southeast thirty-three per cent of the counties of
the state, yet sixty-three per cent of the population.
Thus it can be said that in a remarkable measure the mound-
building peoples found with interesting exactness the portions of this
country which have been the choice spots with the race which now
occupies it.
Here, in the valleys, and between them, toiled their prehistoric
people. Their low grade of civilization is attested by the rude
implements and weapons and domiciles with which they seem to
have been content. Divided, as it is practically sure they were, into
numerous tribes, there must have been some commerce and there
was, undoubtedly, much conflict. Above their poorly cultivated fields,
or in the midst of them, they erected great earthen and stone
fortresses, and, flung far and wide over valley and hill, stand the
mounds in which they buried their dead.

Click here for larger image size

Archaeological Map of Ohio

[Showing interior location of remains]
 CHAPTER III
EARLY TRAVEL IN THE INTERIOR
It has been noted that a considerable portion of archæological
remains in Ohio, Indiana, Michigan, and Wisconsin are inland—or
away from the largest river valleys. The lands on the lesser streams
were occupied in some instances for the entire distance to the
springs. For instance, in Ohio and Kentucky we find only a fraction of
the ancient works on the shores of the Ohio river—either mounds or
forts. In Ohio the largest collections are found in the interior
counties mentioned, as is the case in Kentucky, at a distance varying
from fifty to one hundred and fifty miles from the great water
highway which flows by these states on the north or on the south.
As with these states, so with the counties within them—the
mound-building people seem to have been scattered widely. An
archæological map of Butler county, Ohio, shows that the remains
are found everywhere quite without reference to the largest streams.
In this county there are more works in Oxford township in the far
corner of the county than in Hanover township, which lies between
Oxford and the Miami river. Today there are six hundred more
inhabitants in Oxford township (exclusive of the population of Oxford
village) than in Hanover township. There are more remains in Reily
township, which is separated from the Miami by Ross township, than
in Madison township, which is bounded by the Miami and is drained
by a larger stream than any in Reily. St. Clair township contains
several works in the western portion, on the branches of the Miami
river, and almost none at all in the eastern portion which is bounded
by that river itself.
Crawford county, Wisconsin, has also been mapped. Though
bounded on the south and west by the Wisconsin and Mississippi
rivers, fifty per cent of the ancient works are at a distance from
those streams.

Click here for larger image size

Archaeologic Map of Illinois and Indiana

[Showing interior location of remains]
 The large proportion of remains in Kentucky are in the western
portion of the state situated along the watershed between the Ohio
and Tennessee valleys. In Indiana the great majority of works are in
the eastern tier of counties where there are no streams of
importance.
This makes up a sum of testimony that enables one to say that in
some instances at least the mound-building peoples were largely a
rural people; in some noticeable instances their works are found
more profusely on the smaller streams than on larger ones. In this
they differed in no wise from the red-men who were found living in
these regions mentioned when the whites first came to visit them,
and we might have held to our original line of reasoning to reach
this same conclusion. It might have been shown that the red-men in
Ohio and some of the neighboring states lived more on the smaller
streams than on the larger ones, and then made the deduction that
the mound-building people did the same.
For this was true. The three centers of Indian population in Ohio
were on the smaller streams. The Delawares made their
headquarters on the upper Muskingum; the Wyandots had their
villages on the Sandusky river and bay; the Shawanese were on the
Scioto, and the Miamis on the rivers that have borne their name.
The well-known Indian settlements on the Ohio and on Lake Erie can
almost be counted on the fingers of one hand, while the towns at
Coshocton, Chillicothe, Piqua, Fremont, and Dresden were of
national importance during the era of conquest. Referring to the
location of the Indians of Ohio an early pioneer casually writes:
“Their habitations were at the heads of the principal streams.”[7]
There was almost no exception to the rule.
The explanation of this may be found partly in the great floods
which were, doubtless, more menacing near the larger streams.
While the floods rise perhaps faster today, it is doubtful if they reach
the height that they did in earlier days. Then, at flood-tide, a
thousand forest swamps, licks, pools, and lagoons which do not exist
today added their waters to the river tides. General Butler, who was
on the lower Ohio just after the Revolution, was advised by a
friendly Indian chief to locate Fort Finney high up from the Ohio in
order to be clear of high water. Under the date of October 24, 1785,
he wrote in his Journal: “Capt. George, who had lived below the
mouth of this river [Miami] assured me that all the bank from the
river for five miles did absolutely overflow, and that he had to
remove to the hill at least five miles back, which determined me to
take the present situation.”[8]
Under such circumstances as these it is not surprising that the
Indians preferred the little rivers to the larger ones. The smaller
streams amid their hills did not rise so high, and when they did rise
safe camping spots could be found on high ground not far removed.
What was true of these Indians was probably true of any
antecedent race.
Assuming, then, that the mound-building people lived (in these
states more particularly noticed) on the lesser “inland” streams
where the later Indians were found, there is no question but that
they moved about the country more or less as the Indians
themselves did. Although the former people were more nearly a
stationary people, yet we know that they hunted, and it is not
reasonable to believe that they did not have commercial intercourse.
In fact, from the contents of their mounds, we know they did. We
also know that the various tribes made war upon one another, or at
least were made war upon by some enemy.
All this necessitated highways of travel. Any one who has studied
the West during Indian occupancy does not need to be asked to
remember that travel in the earliest days in the interior was by land
as well as by water. Those making long journeys at propitious
seasons, such as the Iroquois who went southward in war parties,
the Moravians being transferred to Ohio from Pennsylvania, pioneers
en route down the Ohio river to Kentucky, the Wyandots on their
memorable hegira to the Detroit river, used the waterways. But the
main mode of travel for explorers, war parties, pioneer armies and
missionaries seems to have been by the paths which threaded the
forests.[9]
Of the hundreds of Indian forays into Virginia and Kentucky there
is perhaps not one, even those moving down the Scioto and up the
Licking, that used water transportation. In their hunting trips the
canoe was useless except for transporting game and peltry to the
nearest posts, and this was done often on the little Indian ponies.
For long months the lesser streams were ice-bound in the winter;
in the summer, for equally long periods, they must have been nearly
dry, as in the present era of slack-water navigation the larger of
them are frequently very low. Even travel on the Ohio in low-water
months was exasperatingly slow. One pilgrim to Ohio spent ninety
days en route from Killingly, Connecticut, to Marietta, Ohio—thirty-
one of them being spent in getting from Williamsport, Pennsylvania,
down the Monongahela and Ohio to Marietta! The journey from
Connecticut in a cart drawn by oxen to the Monongahela took but
twice the time needed to come down the rivers to Marietta on a
“Kentucky” flatboat![10] With high-water, and going down stream, a
hundred miles a day could be covered.[11]
That the first pioneers into the interior of Ohio, Kentucky, and
Indiana preferred land routes to water is plain to the most casual
reader of the history of the pioneer period. Such great entrepots as
Wellsburg, Ohio, Limestone, Kentucky, and Madison, Indiana (all on
the Ohio), attest the fact that the travel to the interior was by land
routes and not by the smaller rivers.
And so, throughout historic times, one rule has held true in the
region now under survey: that the lesser streams have never been
used to any large degree as routes of travel by the white race, or by
the red race before them. It is thus reasonable to believe that the
earliest people, who so largely inhabited the interior valleys, found
land travel more sure and expeditious than water travel on the little
streams. A great many mound-building people lived by these smaller
streams where so many of their works now stand. That they had
ways of getting about the country goes without saying. In some
instances the earth and stone with which they worked was brought
from a distance. This could not have been accomplished by any
means by water. We know they fought great battles; it is exceedingly
doubtful and all against the lesson taught in times that are historic,
that these armies traveled water routes. True, there were watch
“towers” along the river banks, and the rivers of real size were
undoubtedly the routes of armies—but it has been the opinion of
some archæologists that their enemies came from the north. There
are no rivers flowing from the interior of Ohio, for instance, to Lake
Erie that are even now when dammed of a size sufficient to warrant
us the belief that great armies passed over them.[12] We cannot
imagine a hostile army of power great enough to have necessitated
the building of such a work as Fort Ancient ever coming to it on the
little river on which it stands.
Speaking of the mound-building Indians, MacLean remarks: “In
order to warn the settlements [of mound-builders], where such a
band should approach, it was found necessary to have ... signal
stations. Judging by the primitive methods employed, these wars
must have continued for ages. If the settlements along the two
Miamis and Scioto were overrun at the same time before they had
become weakened, it would have required such an army as only a
civilized or semi-civilized nation could send into the field. It is
plausible to assume that a predatory warfare was carried on at first,
and on account of this the many fortifications were gradually built.
During a warfare such as this, the regular parties of miners would go
to the mines, for the roads could be kept open, even should an
enemy cross the well-beaten paths.”[13] Here a scholar of reputation
gives the strongest kind of evidence in a belief that overland routes
of travel were in existence and were employed in prehistoric times—
by incidentally referring to them while discussing another question.
It is difficult to think of any possible alternative. The verdict of
history is all against another.
Assuming, then, that overland routes of travel were used by this
earliest of American races of which we have any real knowledge, it is
to the purpose of our study to consider where such routes were laid.
The one law which has governed land travel throughout history is
the law of least resistance, or least elevation. “An easy trail to high
ground” is a colloquial expression common in the Far West, but there
has been a time when it was as common to Pennsylvania and Ohio
as it is common today along the great stretches of the Platte. The
watersheds have been the highways and highestways of the world’s
travel. The farther back we go in our history, the more conclusive
does the evidence become that the first ways were the highestways.
Our first roads were ridge roads and their day is not altogether past
in many parts of the land. These first roads were “run,” or built,
along the general alignment of the first pioneer roads, which, in
turn, were nothing more than “blazed” paths of the Indian and
buffalo. A single glance at one of the maps of the Central West of
Revolutionary times, for instance, will show how closely the first
routes clung to the heights of the watersheds. And for good reason:
here the ground suffered least from erosion; here the forests were
thinnest; here a pathway would be swept clear of snow in winter
and of leaves in summer by the swift, clean brooms of the winds.
For the Indians, too, the high lands were points of vantage both in
hunting and in times of war.
In every state there were strategic heights of land, generally
running westward; in Ohio, for example, the strategic watershed was
that between the heads of the lake rivers and the heads of those
flowing southward into the Ohio. Across this divide ran the Great
Trail toward Detroit and the lake country. In western Virginia a
strategic watershed was that formed between the heads of streams
flowing northward into the Ohio and southward into the Kanawha.
And in a remarkable degree the strategic points of a century and a
half ago are the strategic points today, a fact attested by the courses
of the more important trunk railway lines. The steady rise and
importance of such a city as Akron, Ohio, is due to a strategic
situation at the junction of both an important portage path and of a
great watershed highway.
With all these facts in mind it is not presumptuous then to inquire
whether the mound-building Indians did not find the high lands and
mark out on them these first highways of America.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

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

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com