Download the full version and explore a variety of ebooks

or textbooks at https://ebookultra.com

Handbook of Constraint Programming 1st Edition

Francesca Rossi

_____ Tap the link below to start your download _____

https://ebookultra.com/download/handbook-of-constraint-
programming-1st-edition-francesca-rossi/

Find ebooks or textbooks at ebookultra.com today!

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

Francesca 1st Edition Bertrice Small

https://ebookultra.com/download/francesca-1st-edition-bertrice-small/

Handbook of SAS DATA Step Programming 1st Edition Arthur

Li

https://ebookultra.com/download/handbook-of-sas-data-step-
programming-1st-edition-arthur-li/

Constraint Handling Rules 1st Edition Thom Frühwirth

https://ebookultra.com/download/constraint-handling-rules-1st-edition-
thom-fruhwirth/

Philosophical Posthumanism 1st Edition Francesca Ferrando

https://ebookultra.com/download/philosophical-posthumanism-1st-
edition-francesca-ferrando/
Unity in the Book of Isaiah 1st Edition Benedetta Rossi

https://ebookultra.com/download/unity-in-the-book-of-isaiah-1st-
edition-benedetta-rossi/

The Metre of Beowulf A Constraint Based Approach Michael

Getty

https://ebookultra.com/download/the-metre-of-beowulf-a-constraint-
based-approach-michael-getty/

Abstinence Education 1st Edition Isabella E. Rossi

https://ebookultra.com/download/abstinence-education-1st-edition-
isabella-e-rossi/

Supranational Citizenship and the Challenge of Diversity

1st Edition Francesca Strumia

https://ebookultra.com/download/supranational-citizenship-and-the-
challenge-of-diversity-1st-edition-francesca-strumia/

The Linux programming interface a Linux and UNIX system

programming handbook 1st Edition Michael Kerrisk

https://ebookultra.com/download/the-linux-programming-interface-a-
linux-and-unix-system-programming-handbook-1st-edition-michael-
kerrisk/
Handbook of Constraint Programming 1st Edition
Francesca Rossi Digital Instant Download
Author(s): Francesca Rossi, Peter van Beek, Toby Walsh, (eds.)
ISBN(s): 9780444527264, 0444527265
Edition: 1
File Details: PDF, 14.79 MB
Year: 2006
Language: english
 FOUNDATIONS OF
ARTIFICIAL INTELLIGENCE
 Foundations of Artificial Intelligence

Series Editors

J. Hendler
H. Kitano
B. Nebel

Cover picture by Helmut Simonis

ELSEVIER
AMSTERDAM–BOSTON–HEIDELBERG–LONDON–NEW YORK–OXFORD
PARIS–SAN DIEGO–SAN FRANCISCO–SINGAPORE–SYDNEY–TOKYO
 Handbook of Constraint Programming

Edited by

Francesca Rossi
University of Padova
Italy

Peter van Beek

University of Waterloo
Canada

Toby Walsh
National ICTA Australia &
University of New South Wales
Australia

ELSEVIER
AMSTERDAM–BOSTON–HEIDELBERG–LONDON–NEW YORK–OXFORD
PARIS–SAN DIEGO–SAN FRANCISCO–SINGAPORE–SYDNEY–TOKYO
Elsevier
Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK

First edition 2006

Copyright © 2006 Elsevier B.V. All rights reserved

No part of this publication may be reproduced, stored in a retrieval system

or transmitted in any form or by any means electronic, mechanical, photocopying,
recording or otherwise without the prior written permission of the publisher

Permissions may be sought directly from Elsevier’s Science & Technology Rights
Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333;
email: permissions@elsevier.com. Alternatively you can submit your request online by
visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting
Obtaining permission to use Elsevier material

Notice
No responsibility is assumed by the publisher for any injury and/or damage to persons
or property as a matter of products liability, negligence or otherwise, or from any use
or operation of any methods, products, instructions or ideas contained in the material
herein. Because of rapid advances in the medical sciences, in particular, independent
verification of diagnoses and drug dosages should be made

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

ISBN-13: 978-0-444-52726-4
ISBN-10: 0-444-52726-5
ISSN: 1574-6525

For information on all Elsevier publications

visit our website at books.elsevier.com

Printed and bound in The Netherlands

06 07 08 09 10 10 9 8 7 6 5 4 3 2 1
Foreword
Constraints are an ubiquitous concept, which in its broader sense pertains to every day
experience: they represent the conditions which restrict our freedom of decision. In fact,
how much our choices are constrained by the external world is a basic philosophical ques-
tion. In the formalized reasoning of scientific disciplines, constraints have been employed
extensively, from logic to numerical analysis, from mathematical programming to opera-
tions research. In computer science, constraints have been with us from the early days, for
modeling, representing and reasoning (see the interesting historical remarks in Chapter 2
of this handbook, Constraint Satisfaction: An Emerging Paradigm).
I see several good reasons for this ubiquity: one is the conceptually clear separation
between the perfectly declarative problem statements and the often cumbersome enumera-
tive efforts for finding solutions. Another reason is the complexity challenge: the classical
constraint satisfaction problem is NP-complete and in fact tautology checking in propo-
sitional calculus (a constraint problem on Boolean variables) has been the touchstone for
this complexity class. A further reason is that large, complex constraint problems often
occur in practice, they must be solved in one way or another, and fast, efficient, systematic
solutions have an enormous economic value.
What I find surprising about constraints is that within artificial intelligence and com-
puter science a relatively recent, relatively uniform body of knowledge has emerged which
often yields decisive advantages over classical, extensively studied and well developed
techniques. As for many success stories within computer science, success is largely due
to a mixture of structures, algorithms, languages, programming techniques and system im-
plementations. The aim of this handbook is to present this knowledge in all its facets.
Different chapters are largely self contained and all contribute to put the subject into focus,
similarly to the Hawaii Keck observatory, where the mirror is composed of 36 hexagonal
segments.
From the conceptual point of view, the main characteristic features of constraint pro-
gramming are constraint propagation, and the identification of various special cases which
make complexity tractable. The former (see Chapter 3) is an inference technique which
makes local constraints stronger without changing the global constraint. The latter issue
concerns both the structure (see Chapter 7, Tractable Structures for Constraint Satisfaction
Problems) and the kind of constraints (see Chapter 8, The Complexity of Constraint Lan-
guages). Less specific, but still very important issues are as follows: Backtracking Search
Algorithms, in Chapter 4; Local Search, in Chapter 5; Global Constraints, in Chapter 6;
Symmetry in Constraint Programming, in Chapter 10; and Modelling, in Chapter 11.
Another surprising fact about constraint theory is the incredibly close relationship with
logic programming. In a rather precise sense logic programming is a way of expressing,
and solving, certain classes of disjunctive, recursive constraints. Furthermore, logic pro-
gramming can be very elegantly generalized to constraint logic programming (see Chapter

v
vi Foreword

12), where the ordinary Herbrand constraint system, and its unification algorithm, are com-
plemented with specific constraint solvers. The interaction with the committed choice lan-
guages studied in the Japanese projects of the eighties also yielded very interesting models
of computation based on constraints. Amalgamation with more common (and efficiently
implemented!) programming languages is also possible (see Chapter 13, Constraints in
Procedural and Concurrent Languages).
Besides and beyond the beauty of its theoretical foundations, what contributes the most
to the practical convenience of constraint programming are: (i) the development of specific
results for important classes of constraints; (ii) the ability of extending the basic theory to
various additional aspects which are very relevant in practice; and (iii) the flexibility and
potential for integration with other modeling and solving methodologies.
About the development of specific results, this handbook includes chapters about con-
straints on finite (Chapter 14), structured (Chapter 17), temporal (Chapter 19), continuous
and interval-based (Chapter 16) domains. The potential to extend the basic theory in evi-
dent in the case of soft constraints, considered in Chapter 9. Ordinary constraints are either
satisfied or not, namely either true or false. Instead soft constraints return a more infor-
mative weight. Interestingly enough, the proposed extensions both accommodate several
important cases (fuzzy, hierarchical, optimization, probabilistic constraints), and still of-
ten exhibit essentially the same solution algorithms. Extensions to random, changing and
distributed/open constraints are treated in Chapters 18, 21 and 20 respectively.
About the last issue, in addition to the seamless integration with logic and imperative
programming languages we mentioned already, quite remarkable are the paradigms result-
ing from the integration of constraint programming with operations research (see Chapter
15), with scheduling and planning (see Chapter 22), with vehicle routing (see Chapter 23),
with component configuration (see Chapter 24), with (electricity, water, oil, data) networks
(see Chapter 25), and with bioinformatics (see Chapter 26).
The global scenario based on service-oriented computing which is now under devel-
opment offers additional theoretical and practical challenges to constraint programming.
Conditions for service deployment and discovery, both functional and involving different
aspects of quality of service, could be expressed in terms of hard and soft constraints, and
the negotiation phases should involve substantial constraint solving abilities. Transactions
among the various actors could also require partially backtrackable behavior or at least
programmable compensations. Some level of real time, distributed, global constraint solv-
ing should be implemented in the middleware, since lots of higher level applications will
be able to take advantage of, and pay for it.
I think that research and practical development in the area of constraint programming
will be very active for quite a while in the future, establishing closer and closer connections
with a variety of other design methodologies and even other disciplines. I consider this
handbook not only a very nice piece of scientific work, but also a contribution quite instru-
mental at disseminating advanced knowledge about constraint programming both within
the inner constraint community and across the much wider audience of potential users.

U GO M ONTANARI
Dipartimento di Informatica
Università di Pisa, Italy
Editors
Francesca Rossi
University of Padova
Italy

Peter van Beek

University of Waterloo
Canada

Toby Walsh
National ICT Australia &
University of New South Wales
Australia

vii
This page intentionally left blank
Contributors
Rolf Backofen Eugene C. Freuder
Albert-Ludwigs-Universität Cork Constraint Computation Centre &
Germany University College Cork, Ireland

Philippe Baptiste Thom Frühwirth

CNRS LIX & École Polytechnique Universität Ulm
France Germany

Frédéric Benhamou Ian P. Gent

Université de Nantes University of St. Andrews
France Scotland, United Kingdom

Christian Bessiere Carmen Gervet

LIRMM-CNRS Brown University
France USA

Kenneth N. Brown David Gilbert

Cork Constraint Computation Centre & University of Glasgow
University College Cork, Ireland Scotland, United Kingdom

Mats Carlsson Carla Gomes

SICS AB Cornell University
Sweden USA

David Cohen Laurent Granvilliers

Royal Holloway, University of London Université de Nantes
United Kingdom France

Rina Dechter John N. Hooker

University of California, Irvine Carnegie Mellon University
USA USA

Boi Faltings Holger H. Hoos

Swiss Federal Institute of Technology University of British Columbia
Switzerland Canada

ix
x Contributors

Peter Jeavons Wim Nuijten

University of Oxford ILOG SA
United Kingdom France

Ulrich Junker Karen E. Petrie

ILOG SA University of St. Andrews
France Scotland, United Kingdom

Irit Katriel Jean-François Puget

University of Aarhus ILOG SA
Denmark France

Philip Kilby Francesca Rossi

The Australian National University University of Padova
Australia Italy

Manolis Koubarakis Thomas Schiex

University of Athens INRA Toulouse
Greece France

Philippe Laborie Christian Schulte

ILOG SA KTH - Royal Institute of Technology
France Sweden

Claude Le Pape Paul Shaw

ILOG SA ILOG SA
France France

Alan K. Mackworth Helmut Simonis

University of British Columbia CrossCore Optimization
Canada United Kingdom

Kim Marriott Barbara M. Smith

Monash University Cork Constraint Computation Centre &
Australia University College Cork, Ireland

Pedro Meseguer Peter J. Stuckey

IIIA-CSIC University of Melbourne
Spain Australia

Laurent Michel Edward Tsang

University of Connecticut University of Essex
USA United Kingdom

Ian Miguel Peter van Beek

The University of St. Andrews University of Waterloo
Scotland, United Kingdom Canada
 Contributors xi

Willem-Jan van Hoeve

Cornell University
USA

Mark Wallace
Monash University
Australia

Toby Walsh
National ICT Australia &
University of New South Wales
Australia
This page intentionally left blank
Contents

Foreword v

Editors vii

Contributors ix

Contents xiii

I Foundations 1

1 Introduction 3
Francesca Rossi, Peter van Beek, Toby Walsh
1.1 Purpose of the Handbook . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 Structure and Content . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2 Constraint Satisfaction: An Emerging Paradigm 13

Eugene C. Freuder and Alan K. Mackworth
2.1 The Early Days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2 The Constraint Satisfaction Problem: Representation and Reasoning . . 16
2.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

3 Constraint Propagation 29
Christian Bessiere
3.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.2 Formal Viewpoint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.3 Arc Consistency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.4 Higher Order Consistencies . . . . . . . . . . . . . . . . . . . . . . . . 50
3.5 Domain-Based Consistencies Stronger than AC . . . . . . . . . . . . . 57
3.6 Domain-Based Consistencies Weaker than AC . . . . . . . . . . . . . . 62
3.7 Constraint Propagation as Iteration of Reduction Rules . . . . . . . . . . 68
3.8 Specific Constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

4 Backtracking Search Algorithms 85

Peter van Beek
4.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.2 Branching Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

xiii
xiv Contents

4.3 Constraint Propagation . . . . . . . . . . . . . . . . . . . . . . . . . . 90

4.4 Nogood Recording . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
4.5 Non-Chronological Backtracking . . . . . . . . . . . . . . . . . . . . . 102
4.6 Heuristics for Backtracking Algorithms . . . . . . . . . . . . . . . . . . 105
4.7 Randomization and Restart Strategies . . . . . . . . . . . . . . . . . . . 111
4.8 Best-First Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
4.9 Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
4.10 Comparing Backtracking Algorithms . . . . . . . . . . . . . . . . . . . 118

5 Local Search Methods 135

Holger H. Hoos and Edward Tsang
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
5.2 Randomised Iterative Improvement Algorithms . . . . . . . . . . . . . 142
5.3 Tabu Search and Related Algorithms . . . . . . . . . . . . . . . . . . . 144
5.4 Penalty-Based Local Search Algorithms . . . . . . . . . . . . . . . . . 148
5.5 Other Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
5.6 Local Search for Constraint Optimisation Problems . . . . . . . . . . . 155
5.7 Frameworks and Toolkits for Local Search . . . . . . . . . . . . . . . . 157
5.8 Conclusions and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . 158

6 Global Constraints 169

Willem-Jan van Hoeve and Irit Katriel
6.1 Notation and Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . 170
6.2 Examples of Global Constraints . . . . . . . . . . . . . . . . . . . . . . 176
6.3 Complete Filtering Algorithms . . . . . . . . . . . . . . . . . . . . . . 182
6.4 Optimization Constraints . . . . . . . . . . . . . . . . . . . . . . . . . 189
6.5 Partial Filtering Algorithms . . . . . . . . . . . . . . . . . . . . . . . . 193
6.6 Global Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
6.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203

7 Tractable Structures for Constraint Satisfaction Problems 209

Rina Dechter
7.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
7.2 Structure-Based Tractability in Inference . . . . . . . . . . . . . . . . . 213
7.3 Trading Time and Space by Hybrids of Search and Inference . . . . . . 231
7.4 Structure-Based Tractability in Search . . . . . . . . . . . . . . . . . . 239
7.5 Summary and Bibliographical Notes . . . . . . . . . . . . . . . . . . . 241

8 The Complexity of Constraint Languages 245

David Cohen and Peter Jeavons
8.1 Basic Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
8.2 Examples of Constraint Languages . . . . . . . . . . . . . . . . . . . . 247
8.3 Developing an Algebraic Theory . . . . . . . . . . . . . . . . . . . . . 251
8.4 Applications of the Algebraic Theory . . . . . . . . . . . . . . . . . . . 258
8.5 Constraint Languages Over an Infinite Set . . . . . . . . . . . . . . . . 263
8.6 Multi-Sorted Constraint Languages . . . . . . . . . . . . . . . . . . . . 264
8.7 Alternative Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . 269
Exploring the Variety of Random
Documents with Different Content
The Project Gutenberg eBook of Nitrogen
Bacteria and Legumes
This ebook is for the use of anyone anywhere in the United
States and most other parts of the world at no cost and with
almost no restrictions whatsoever. You may copy it, give it away
or re-use it under the terms of the Project Gutenberg License
included with this ebook or online at www.gutenberg.org. If you
are not located in the United States, you will have to check the
laws of the country where you are located before using this
eBook.

Title: Nitrogen Bacteria and Legumes

Author: Cyril G. Hopkins

Release date: January 31, 2018 [eBook #56472]

Language: English

Credits: Produced by Bryan Ness, Jana Srna and the Online

Distributed
Proofreading Team at http://www.pgdp.net (This file
was
produced from images generously made available by
The
Internet Archive/American Libraries.)

*** START OF THE PROJECT GUTENBERG EBOOK NITROGEN

BACTERIA AND LEGUMES ***
 UNIVERSITY OF ILLINOIS
Agricultural Experiment Station.
BULLETIN NO. 94.

NITROGEN BACTERIA AND

LEGUMES
(With Special Reference to Red Clover, Cowpeas, Soy Beans, Alfalfa, and Sweet Clover,
on Illinois Soils).

By CYRIL G. HOPKINS.

Work and Knowledge are a stronger team than Work and Work.

URBANA, ILLINOIS, FEBRUARY, 1904.

 Summary of Bulletin No. 94.
1. Soil nitrogen cannot be used by plants until it is
changed to the form of nitrate nitrogen by the
nitrifying bacteria. Page 307
2. Atmospheric nitrogen cannot be used by any
agricultural plants, excepting legumes, and even
leguminous plants have no power to obtain nitrogen
from the air unless they are provided with the proper
nitrogen-gathering bacteria. Page 309
3. As a rule each important agricultural legume must
have its own particular species of bacteria. Page 311
4. The frequent failure of red clover in normal
seasons, especially on normal soils occupying the
highest land, is undoubtedly due in part at least to the
absence of the proper bacteria (sometimes the soil
lacks lime or phosphorus). Page 313
5. On the very acid soils, where clover has never
been grown successfully, applications of ground
limestone should be made where legumes are to be
grown. Page 315
6. Cowpeas need not be inoculated, because the
cowpea bacteria are usually either present in the soil
or are introduced with the seed in sufficient numbers
to effect a good degree of infection if the soil is
suitable and if cowpeas are seeded upon the same
land for two successive years. Page 317
7. Cowpeas grown on infected soil on the University
of Illinois soil experiment field contained four times as
much nitrogen as the same kind of cowpeas grown on
similar land which was not infected. Page 319
 8. As a rule soy bean fields should be inoculated
when first seeded to soy beans, otherwise they may be
grown on the same land for three or four years before
the soil becomes thoroughly infected. Page 320
9. Investigations, reported in this bulletin, furnish
conclusive proof that infected sweet clover soil can be
used for inoculating alfalfa fields and with the same
results as are obtained when the infected soil is
obtained from an old alfalfa field. (Sweet clover is a
tall, rank-growing, sweet-scented leguminous plant
widely distributed over Illinois, especially along the
roadsides in the northern and central parts of the
state, and in many places in bottom lands in southern
Illinois.) Page 324
10. This bulletin will be sent free of charge to any
one interested in Illinois agriculture upon request to E.
Davenport, Director Agricultural Experiment Station,
Urbana, Illinois; and, if so requested, the name of the
applicant will be placed upon the permanent mailing
list of the Experiment Station, so that all bulletins will
be sent to him as they are issued.
Conclusions Page 327
 NITROGEN BACTERIA AND
LEGUMES
(With Special Reference to Red Clover,
Cowpeas, Soy Beans, Alfalfa, and Sweet
Clover, on Illinois Soils).
By CYRIL G. HOPKINS, Chief in Agronomy and Chemistry.
Among the several different classes or groups of bacteria there are
two which are of special importance to agriculture because of their
relation to the element nitrogen, this being commonly considered
the most valuable element of plant food.[1] These two classes of
bacteria are, first, the nitrifying bacteria, and, second, the nitrogen-
gathering bacteria.

The Nitrifying Bacteria.

The nitrifying bacteria are those which have the power to form
nitrates. In the following brief discussion of this subject we include
at least three species of bacteria which by their combined or
successive action have the power to transform organic nitrogen into
nitrate nitrogen, which is a suitable form of nitrogen for plant food.
For the exact information which we now have regarding the
nitrifying bacteria we are indebted to the researches of Pasteur and
Schlösing and Müntz of France, Winogradsky of Russia, Warington of
England, and others.
 The nitrogen in the soil is almost entirely in organic compounds;
that is, the nitrogen (which is a gas in the free, or uncombined,
state) is united or combined with other elements, notably with
carbon, hydrogen, and oxygen, in the form of partially decayed
vegetable or organic matter. (By organic matter we mean matter
which has been formed by the growth of some organism, either
plant or animal, as grass or flesh.) Plants cannot use the free
nitrogen of the air as plant food, neither can they use the organic
compounds of nitrogen which occur in the soil. There are at least
three different kinds of bacteria, and also three different steps or
stages involved in the process of nitrification, the nitrogen being
changed from the organic compounds first into the ammonia[2]
form, second, into the nitrite form, and third into the nitrate form.
During the process the nitrogen is separated from the carbon and
other elements composing the insoluble organic matter, and is united
or combined with oxygen and some alkaline element (as calcium) to
form the soluble nitrate, such as calcium nitrate, which is one of the
most suitable compounds of nitrogen for plant food. Calcium is the
alkaline element contained in lime or limestone. The name calcium
nitrate indicates just what elements this compound contains;
namely, calcium, nitrogen, and oxygen. (In the names of compounds
the ending -ate always means oxygen.)
This is the general process of nitrification in which the nitrifying
bacteria transform or transfer the nitrogen from insoluble organic
compounds into soluble compounds in which it may serve as
available plant food. The nitrate which is thus formed may be
calcium nitrate or magnesium nitrate or potassium nitrate or even
sodium nitrate, depending upon which of these alkaline elements is
present in the must suitable form. If no alkaline element is present
in available form then no nitrates can be made in the soil. One of the
reasons for applying ground limestone to soils which are deficient in
lime is to furnish the element calcium in suitable form for the
formation of nitrates in the process of nitrification. Ground limestone
is calcium carbonate (CaCO₃), a compound containing one atom of
calcium (Ca), one atom of carbon (C) and three atoms of oxygen
(O₃). This is the same form of lime which is contained naturally in
limestone soils—soils which are noted for their great productiveness
—and it is generally the most economical form of lime to use for
correcting soil acidity and promoting nitrification.
In the process of nitrification, that is in the formation of nitrates,
there is required, not only the presence of calcium, or some other
alkaline element, in suitable form, but also a good supply of the
element oxygen; for calcium nitrate, Ca(NO₃)₂, contains one atom of
calcium (Ca), two atoms of nitrogen (N)₂, and six atoms of oxygen
(O₃)₂, in each molecule as indicated in the formula, Ca(NO₃)₂.
Magnesium nitrate, Mg(NO₃)₂, potassium nitrate, KNO₃ (K is from
the Latin word Kalium, which means potassium), and all other
nitrates, also, contain oxygen. The supply of oxygen for the
formation of nitrates in the soil comes from the air, which consists of
about twenty percent oxygen, seventy-eight percent nitrogen, and
two percent of other elements and compounds, as argon, carbon
dioxid, CO₂, water vapor, H₂O, etc. One of the important effects of
cultivation, or tillage, is that it permits the air more freely to enter
the soil, and thus promotes nitrification.

The Nitrogen-Gathering Bacteria.

As stated above, the nitrogen naturally in the soil is contained
almost entirely in the organic matter. Any process which tends to
decompose or destroy this organic matter, such as nitrification or
other forms of oxidation, will also tend to reduce the total stock of
nitrogen in the soil. Because of this fact the matter of restoring
nitrogen to the soil becomes of very great importance. Of course a
part of the nitrogen removed in crops may be returned in the
manure produced on the farm; and nitrogen may also be bought in
the markets in such forms as sodium nitrate (containing 15 to 16
percent of nitrogen), ammonium sulfate (containing 20 to 21 percent
of nitrogen), and dried blood (containing 12 to 15 percent nitrogen);
but, when we bear in mind that such commercial nitrogen costs
about 15 cents a pound, and that one bushel of corn contains about
one pound of nitrogen, it will be seen at once that the purchase of
nitrogen cannot be considered practicable in general farming,
although in market gardening, and in some other kinds of intensive
agriculture, commercial nitrogen can often he used with very marked
profit.
Nitrogen is removed from the soil not only in the crops grown, but
also, and frequently in larger amounts per annum, in the drainage
waters, and in some other ways, as by denitrification and by the
blowing and washing of the surface soil. Professor Snyder, of the
Minnesota Experiment Station, has shown that during a series of
years the total loss of nitrogen from some Minnesota soils in some
cases amounts to several times the amount actually used in the
crops produced.
Considering all of these facts, and the additional facts that there
are about seventy-five million pounds of atmospheric nitrogen
resting upon every acre of land, and that it is possible to obtain
unlimited quantities of nitrogen from the air for use of farm crops,
and at very small cost, the inevitable conclusion is that the
inexhaustible supply of nitrogen in the air is the store from which we
must draw to maintain a sufficient amount of this element in the soil
for the most profitable crop yields.
It is often stated that leguminous plants, such as clover, have
power to obtain free nitrogen from the air. This is not strictly true.
Red clover, for example, has no power in itself to get nitrogen from
the air. It is true, however, that the microscopic organisms[3] which
commonly live in tubercles upon the roots of the clover plant do
have the power to take free nitrogen from the air and cause it to
unite with other elements to form compounds suitable for plant
food. The clover plant then draws upon this combined nitrogen in
the root tubercles, and makes use of it in its own growth, both in the
tops and in the roots of the plant.
 These nitrogen-gathering bacteria live in tubercles upon the roots
of various leguminous plants,[4] such as red clover, white clover,
alfalfa, sweet clover, cowpeas, soy beans, vetch, field-peas, garden-
peas, field and garden beans, etc. These tubercles vary in size from
a pinhead to a pea, varying with the different kinds of plants, being
especially small upon some of the clovers, and very large upon
cowpeas and soy beans. The tubercles are, of course, easily seen
with the eye, but the tubercle is only the home of the bacteria,
somewhat as the ball upon the willow twig is the home of the
insects within. The bacteria themselves are far too small to be seen
with the unaided eye, although they can be seen by means of the
most powerful microscope. Several million bacteria may inhabit a
single tubercle. It is not necessary to see the bacteria, because if we
find the tubercles upon the roots of the plant, we know that the
bacteria are present within, as otherwise the tubercle would not be
formed.
Although the plant itself, as clover, for example, has no power to
feed upon the free or uncombined nitrogen in the air, yet these
nitrogen-gathering bacteria do have the power to absorb the free
nitrogen and cause it to combine with other elements, forming
nitrates or other compounds which are suitable forms of nitrogen for
plant food.
It has also been demonstrated that, as a rule, there are different
species of nitrogen-gathering bacteria for markedly different species
of leguminous plants. Thus we have one kind of bacteria for red
clover, another kind for cowpeas, another kind for soy beans, and
still a different kind for alfalfa.[5]

The Red Clover Bacteria.

 That clover has no power in itself to gather atmospheric nitrogen,
and that the bacteria do have power to feed the clover plant with
nitrogen gathered from the air is very easy to demonstrate. It is one
of the regular laboratory practices of the students in soil fertility in
the Agricultural College to make this demonstration. Plate 1 is an
illustration of such student work. The two pots which are shown
were provided with all elements of plant food, excepting the one
element nitrogen. Thus far the two pots are exactly alike. Each
contains no nitrogen, as indicated by the label “No N.” Each pot is
planted with the same number of red clover seeds. To the right-hand
pot, however, some bacteria (“Bac.”) were added, while none were
added to the left-hand pot. These bacteria were obtained by taking
about one pound of soil from a clover-field where abundance of
tubercles were found on the clover roots, adding this soil to about
one quart of pure water, shaking for a few minutes, allowing the soil
to settle, then taking a small quantity of almost clear solution, and
adding it to the pot which we wished to inoculate with the red clover
bacteria. Aside from the addition of these microscopic bacteria to the
right-hand pot, these two pots were treated exactly alike throughout
the experiment. It will be plainly seen that where the bacteria were
added the clover was furnished with sufficient nitrogen to make a
strong and luxuriant growth, while without the bacteria the clover (in
the left-hand pot) only germinated and made what little growth it
could with the small amount of nitrogen contained in the seed. This
result is the difference between success and failure of the clover
crop.
 Plate 1. Red Clover: Effect of Bacteria. No Nitrogen in
the Soil of Either Pot.

In general the clover bacteria are well distributed over the

northern and central part of Illinois, but we now have some very
strong evidence that they are not well distributed in some soils of
large area in southern Illinois. There is also some evidence that they
were not originally present even in the soils where they are now
found in great abundance; and, furthermore, it seems very probable
that these bacteria may cease to live in a soil where they have once
been present, provided clover is not grown on the land for several
years.
It will help us to understand this matter if we bear in mind that
the home of these bacteria is the tubercle upon the clover root. It is
quite evident that they will continue to live upon the decaying
tubercles or roots for three or four years after the clover plant has
been killed. On the other hand, we have some notable evidence that
the bacteria do not continue to live in a soil after five or six years’
continuous cropping with absolutely no clover growing on the land
during those years. It is a simple matter for any one to determine
whether the bacteria are present or not, for the tubercles which are
formed if the bacteria are present are plainly seen attached to small
roots. They look somewhat like miniature potatoes, varying in size
from pinheads on clover to peas on soy beans or cowpeas. (See
Plates 2 and 4.) It is important to remember that the bacteria live in
the soil and not in the seed.
When clover is cut for seed, it is frequently left to lie upon the
ground until the straw becomes half rotten and very dirty; and,
consequently when it is threshed, it practically always happens that
there is at least some small amount of dust and dirt taken with the
seed. This dirt is almost sure to carry with it some bacteria from the
soil. If these few bacteria are scattered with the clover seed when it
is sowed they will inoculate at least a few plants, and if they are
allowed to multiply on these plants, and especially if the same field
is repeatedly seeded with clover, the soil will ultimately become
thoroughly infected with the clover bacteria. Of course they may be
carried from one part of the farm to another, or even from one farm
to another, by various agencies, as dust or wind storms, surface
drainage or flood waters, manure made from clover hay, implements
used in cultivating the soil, etc., etc.
Many of the older farmers of Illinois have stated to the writer that
when this country was very new it was commonly found difficult to
get a “catch” of clover on new land. After a good “catch” was once
gotten, then it was easier to get clover to grow on that land the next
time. There was a saying among the farmers that clover would not
do well until they got the “wild nature” out of the land. Their final
success was undoubtedly due, not to getting anything out of the
land, but rather to getting the bacteria into the land. Several Illinois
farmers have reported some quite remarkable results from very light
applications of the clover chaff or straw (obtained in hulling clover)
in its beneficial effect on clover on land where it was otherwise
difficult to get a “catch.” There is a somewhat general belief among
farmers of long experience that clover straw or chaff has some
special value in getting a catch of clover aside from its value as
manure or for the seed which it sometimes contains.
Manager F. A. Warner of the Sibley Estate, Ford County, recently
stated to the writer that they had had very great difficulty to get
clover to grow when they first began growing clover on that large
estate, some six or eight years ago, although, after a good crop was
once secured, they rarely had any further difficulty in getting a catch
of clover on the same land.
On the common gray prairie soil of the Lower Illinoisan Glaciation,
in southern Illinois, the commonest type of soil in more than twenty
counties, practically no red clover is grown. In the spring of 1903 we
seeded red clover on that type of soil in three places; namely, on the
University of Illinois soil experiment fields near Edgewood, Effingham
County, near Du Bois, Washington County, and near Cutler, Perry
County. On certain plots the soil acidity had been corrected with lime
and an abundant supply of phosphorus (in bone meal) had been
provided, potassium also having been added on some plots. These
fields were carefully examined the latter part of June, and at Cutler
and Du Bois the clover was found to be dead or dying, and no
tubercles could be found upon the clover roots, although on the
clover which had been seeded at about the same time on the
University fields at Urbana the root tubercles were found in great
abundance. At Edgewood a few tubercles were found and the clover
appeared to be growing fairly well. Infected red clover soil was at
once procured and scattered over the fields at Edgewood, Du Bois,
and Cutler, but it was evidently too late to be of any marked benefit.
At Cutler and at Du Bois the clover was a complete failure. (It will be
tried again next year.) At Edgewood it continued to grow fairly well,
and its progress next season (1904) will be watched with much
interest. It should be stated that the Experiment Station has been
growing clover for several years with varying degrees of failure on
land adjoining the present clover field at Edgewood, and it is
possible that this year’s apparent success from the start is due in
part at least to the bacteria which have been incidentally introduced
and multiplied year after year and scattered over the adjoining land
by wind and dust storms. Before the close of the season the
tubercles developed in abundance on the roots of the clover at
Edgewood.
An experience reported by Professor Herbert W. Mumford, of the
Animal Husbandry Department of this university, will be of interest
and value in this connection. Professor Mumford commonly grows
clover in his rotations on his own private farm, but he states that at
one time one particular field was cropped continuously with timothy,
oats, and corn for some six years or more without any clover
whatever. It was then again seeded to clover, but the crop made a
complete failure, although on other land where clover had been
grown more recently a successful clover crop was grown from the
same kind of seed seeded at about the same time. The following
year this particular field was again seeded to clover. This time the
“catch” was not a total failure, but it was too poor to save, and it
was plowed up and the land again seeded to clover the next year,
and an excellent catch of clover resulted. After this, clover was
frequently grown on this field, and no special difficulty was had in
getting good crops.
While the failure of clover may often be due to drouth, and in
some places due to soil acidity (lack of lime), and sometimes even
due to an insufficient supply of available phosphorus or of
potassium, we now know with certainty that it sometimes fails
because of the absence of the nitrogen-gathering bacteria, especially
on land which has never grown clover, and probably also on land
which has not grown it recently. We should always remember that
the bacteria do not thrive in strongly acid soils. Even though they
may sometimes live in such soils and perhaps produce some
tubercles upon the roots of certain hardy, strong growing legumes,
like cowpeas, nevertheless we are obtaining some strong evidence
that in such acid soils they have but little power to gather nitrogen
from the air. That ground limestone is the most economical and
satisfactory material to use in correcting the acidity of soils is
strongly indicated by the information we have thus far obtained. On
the upland prairie soils of the Lower Illinoisan Glaciation where red
clover has never been grown successfully, largely because of the
acidity of the soil, it will undoubtedly be helpful and profitable not
only to correct the acidity of the soil with ground limestone, but also
to secure infected soil from some field of timber land or bottom land
where red clover is growing, well provided with root tubercles, and
inoculate the field with it. This soil should be collected to a depth of
three or four inches and scattered over the prairie land at the rate of
a few hundred pounds per acre at the time the clover is seeded or
before.

The Cowpea Bacteria.

Plate 2 is made from a photograph of a cowpea root with the
tubercles upon it. This illustration shows the cowpea tubercles at
nearly natural size, which is about as large as the seed of ordinary
garden peas.
 Plate 2. Cowpea Root Tubercles, Natural Size.

The cowpea bacteria are already quite widely distributed in

southern Illinois, especially where this crop has been grown for
several years, but they are not common in the soils of other parts of
the state. It is doubtful, however, if it is necessary or even worth
while to take the trouble to inoculate soil for cowpeas. Some few
tubercles almost invariably develop on cowpea roots the first year
they are seeded, even where they have never been grown before,
and if seeded the second year on the same land the plants are
usually abundantly provided with root tubercles. Just why the
cowpea bacteria develop so rapidly even without special inoculation
is not definitely known. It may be that the same bacteria also live on
some other leguminous plant which is more or less widely distributed
over the state, but it seems more likely that the bacteria are brought
with the seed. As a matter of fact, the cowpea harvest is usually
dirty. This is an annual plant, and consequently the crop is grown on
recently plowed land and is sometimes cultivated during the season.
Cowpeas are commonly harvested with a mowing machine and then
raked up on the loose ground. When they are threshed more or less
dirt remains with the seed. Furthermore, the seed coats are not
infrequently cracked, thus providing an excellent place for the
lodgment of particles of soil.
Whether it would be profitable to inoculate the land for cowpeas
would depend very largely upon the difficulty or cost of obtaining the
infected material. If soil thoroughly infected with the cowpea
bacteria can be scattered over the land at the rate of about 2,000
pounds to the acre at a cost of $1.00 or less per ton, it might prove
profitable. It is doubtful if a light application of 100 or 200 pounds
would produce any very marked effect in the yield the first season.
After the soil becomes well infected the cowpeas then obtain much
nitrogen from the air, and the yield of cowpeas is likely to be largely
increased. Of course there is no fixation of atmospheric nitrogen if
there are no tubercles on the roots.
In 1902 several plots of cowpeas were seeded on the soil
experiment field at the university. One of those plots (404) had
become thoroughly infected with the cowpea bacteria because of its
being so situated that more or less surface drainage water flowed
over it from an adjacent field upon which cowpeas had been grown
for three successive years. Another plot (408), owing to a slightly
different situation, had not become infected. The two plots were
seeded in July after a crop of oats had been removed from the land.
Within three weeks after seeding, numerous root tubercles could be
found on the plants on the infected plot. Later on, ten average
consecutive plants were taken up as completely as possible, and 412
tubercles were found on the roots, making an average of more than
40 tubercles to the plant. On Plot 408 only an occasional plant was
found infected, and such plants would usually have only a single
large tubercle on their roots. Ten average plants not infected were
collected from Plot 408 for comparison with the ten infected plants
from Plot 404.

Plate 3. Cowpeas; Effect of Bacteria in Ordinary Illinois

Black Prairie Soil.

Plate 3 shows these two bunches of plants, the infected plants

with root tubercles on the right, and the plants without tubercles on
the left. Four more sets of ten plants each were then collected, two
sets from Plot 404 and two from Plot 408. Each set of infected plants
was separated into three parts, (1) tops, (2) roots, (3) tubercles;
and each set of plants not infected was separated into (1) tops, and
(2) roots. All of these samples were dried and analyzed for nitrogen.
The results obtained are shown in Table 1.
 The results clearly show the very great value of the nitrogen-
gathering bacteria in growing cowpeas. In each of the separate trials
A, B, and C, the infected plants contained about twice as much total
dry matter as the plants not infected. The infected plants also
contained a much higher percent of nitrogen than the plants not
infected, the infected plants containing 4.09 to 4.33 percent in the
tops and 1.45 to 1.53 percent in the roots, while those not infected
contained only 2.32 to 2.69 percent in the tops and .88 percent in
the roots. Besides this, the tubercles on the infected plants contain
5.76 to 6.05 percent of nitrogen. In these young and rapidly growing
plants the tubercles are much richer in nitrogen than any other part
of the plant. It should be stated that as the plants approach maturity
the nitrogen is largely absorbed from the tubercles and stored in the
tops and roots. At the time these plants were taken up the tubercles
actually contained more nitrogen than the roots. The infected plants
contained nearly four times as much nitrogen as the plants not
infected, and about three-fourths of the total nitrogen in the infected
plants was obtained from the air. The roots and tubercles of the
infected plants contained six to seven times as much nitrogen as the
roots of the plants not infected.

Table I.—Fixation of Nitrogen by Cowpeas.

Nitrogen
Cowpea Plants. Dry Nitrogen Nitrogen
fixed by
matter, content, amount,
bacteria,
No. Part. cgs. percent. cgs.
cgs.
A1 Ten Tops 3580 4.09 146 125
— plants, Roots 620 1.45 9
with
Tubercles 190 5.97 11
bacteria
present. Total 4390 166
A2 Ten Tops 1560 2.42 38
— plants, Roots 300 .88 3
 without Total 1860 41
bacteria
B1 Ten Tops 3970 4.31 171
— plants, Roots 690 1.47 10
with
Tubercles 300 6.05 18
bacteria
present. Total 4960 199 140
B2 Ten Tops 2060 2.69 55
— plants, Roots 430 .88 4
without
bacteria Total 2490 59
C1 Ten Tops 3300 4.33 143
— plants, Roots 520 1.53 8
with
Tubercles 290 5.76 17
bacteria
present. Total 4110 168 124
C2 Ten Tops 1730 2.32 40
— plants, Roots 400 .88 4
without
bacteria Total 2130 44

The Soy Bean Bacteria.

Soy bean bacteria are evidently much less likely to be carried with
the seed than are the cowpea bacteria. The soy bean plant grows
more erectly than the cowpea (see Circular No. 69, “The Cowpea
and Soy Bean in Illinois”), and the crop is quite commonly harvested
with a self-binder which keeps it quite free from dirt. The soy bean
seed is nearly round and smooth, and the seed coat is not commonly
cracked. These facts may explain why the soy bean seed carry so
few bacteria as compared with cowpeas.
 On one of the soil experiment fields on the university farm at
Urbana, where soy beams have been grown for three years, no
tubercles could be found on the plants either the first or second
year, and only an occasional plant with tubercles could be found the
third year. In 1902 a series of plots, some of which had been treated
in different ways with applications of limestone, phosphorus, and
potassium, were seeded with soy beans. No tubercles could be
found at any time during the season on the soy beans growing on
any of the different plots. In 1903 the same plots were again seeded
to soy beans, and at the same time part of each plot was inoculated
with infected soy bean soil drilled in with the seed at the rate of
about 500 pounds of infected soil to the acre. When the plants were
only a few weeks old tubercles were to be found upon many plants
growing where the infected soil had been applied, and before the
close of the season at least half of these plants in the inoculated part
of the field had one or more tubercles upon their roots, and some
plants could be found whose roots were abundantly provided with
tubercles. (See Plate 4.)
On the uninoculated part of the field soy bean plants were
examined probably fifty times during the season, several plants
being taken up each time, but not a single tubercle was found at any
time, notwithstanding that this was the second crop of soy beans
upon this soil. Of course the inoculated part of the field did not
become sufficiently infected to markedly benefit the 1903 crop, but it
is planned to grow soy beans upon this field again in 1904 when the
bacteria will doubtless have multiplied sufficiently to produce marked
results in the growth of the crop.
From these and from other somewhat similar experiments it is
concluded that as a rule soy beans should be inoculated when they
are first seeded, and that they should then be grown a second year
upon the same land. If soy beans are afterward grown upon this
land once in every three or four years, the soil will doubtless remain
well infected with the soy bean bacteria.
 It is believed that 100 pounds of infected soy bean soil per acre
will be sufficient to produce a thorough infection the second year,
and it is improbable that one ton of infected soil per acre would
produce a thorough inoculation the first season. One ton is only
twenty times 100 pounds, while one tubercle which will be produced
during a single season from a single bacterium may contain many
million bacteria, thus it will be seen that it will be more economical
to inoculate rather lightly and allow the bacteria to multiply
themselves rather than to inoculate heavily at great expense.
 Plate 4. Soy Bean Root Tubercles, Natural Size.

It may be stated that the infected soy bean soil used in these
experiments was obtained from Mr A. A. Hinkley of Du Bois, Illinois,
who has been growing soy beans on the same land for many years
until it has become well infected. Mr. Hinkley has consented to
furnish infected soy bean soil so far as he is able to do without
serious interference with his regular work, to any one who may
desire it, at a price which will cover his expense and loss. This will
probably amount to about $1.00 for the first 100 pounds and fifty
cents for each additional 100 pounds, in the shipment, including the
cost of bags, the purchaser to pay freight from Bois station, which is
located in Washington County, Illinois, on the Illinois Central
Railroad.

The Alfalfa and Sweet Clover Bacteria.

That soil inoculation with alfalfa bacteria is commonly of very
great value in growing alfalfa has been shown very conclusively by
the investigations reported in Bulletin No. 76,[6] “Alfalfa on Illinois
Soil.” In some places, however, inoculation was found to be
unnecessary. A careful and extensive investigation of alfalfa growing
in different parts of Illinois revealed the fact, as stated in Bulletin 76,
“that the alfalfa bacteria are certainly present in some places in the
state while in most other places they are certainly not present in
sufficient number to become of appreciable assistance to the alfalfa
within three or four years, and the question naturally arises how it
happens that some fields are already infected while others are not.”
It was suggested in that bulletin that the alfalfa bacteria may “live
on some other plants besides alfalfa and that one of these plants is
native or has been introduced in certain sections” of the state. It
was also suggested “that a few bacteria are always carried with
alfalfa seed, and that if the alfalfa is grown continuously or
repeatedly in any place the soil will finally become thoroughly
infected, and the bacteria will then be carried by flood waters, dust
storms, etc., over adjoining fields, and possibly for long distances,
especially along river valleys.” This latter suggestion was known to
be a fact at the time it was written; and subsequent investigations
have furnished conclusive proof that the alfalfa bacteria do live upon
another plant; namely, the ordinary sweet clover (melilotus alba).
This is a rank-growing leguminous plant, frequently reaching a
height of four to six feet. When young it markedly resembles alfalfa,
but it can easily be distinguished by its characteristic odor when cut
or bruised, as by rubbing between the hands. As the sweet clover
approaches maturity it differs very much from alfalfa. The sweet
clover grows very tall, and usually branches from a main stem. It
has white flowers (there is also a less common yellow variety), and
the seeds are borne in small round pods (usually containing only one
or two seeds each), arranged on long slender spikes, each spike
bearing many pods. The alfalfa commonly grows about two and a
half feet high, with many stems growing from the crown of the root,
especially after it is two or three years old. It bears purple flowers
and peculiar spiral-shaped seed pods. Sweet clover is a biennial
plant, dying after reaching maturity, which commonly occurs the
second year of its growth. Like many other biennial plants, it
probably often lives more than two years if not allowed to produce
seed. Alfalfa is a perennial plant, and it is said that there are alfalfa
fields which have been cut annually for more than fifty years without
reseeding. The similarity of alfalfa and sweet clover when young,
and also the similarity of the tubercles formed on the roots of each
have long been noticed, and the possibility of the same bacteria
living upon both plants has already been suggested in the
agricultural press.
During the season of 1903 the writer spent some time in the
northern part of Illinois in connection with the general and detail
surveys of Illinois soil. Many new fields of alfalfa were observed, and
they were carefully examined for root tubercles. In Winnebago
County, where sweet clover is very prevalent along roadsides and in
waste places, it was noted that the abundance of root tubercles on
the alfalfa plants seemed to be closely related to the presence of
sweet clover in the vicinity, strongly indicating that the bacteria
which live upon sweet clover were also at home upon the alfalfa
roots. These indications were strengthened by further investigations
in Lake County, especially upon the Fowler farm, near Lake Villa,
where a field of alfalfa seeded last spring without artificial
inoculation was found to be thoroughly infected with the bacteria,
and growing vigorously with a good dark green color. This field had
a few sweet clover plants growing in it, and the borders of the field
were covered with sweet clover. Other fields of alfalfa seeded in the
neighborhood at the same time, but upon soils where sweet clover
had not grown near by, were apparently complete failures, many of
the plants having died and most of those still living being only a few
inches high, very weak, and yellow or pale green in color.[7]
In order to obtain more absolute knowledge regarding this
important subject, a series of pot culture experiments has been
carried on under controlled conditions in the pot culture laboratory
at the university. Five pots were filled with sterilized sand which was
practically devoid of plant food. A supply of phosphorus, potassium,
and all other mineral elements necessary for the growth of plants
was added to each of the five pots, care being taken to keep the
sand practically free of combined nitrogen. Alfalfa seed were then
planted in each of the five pots, and at the same time four of the
five pots were inoculated as follows:
Pot No. 1.—Not inoculated (check pot).
Pot No. 2.—Inoculated with bacteria obtained from infected alfalfa
soil.
Pot No. 3.—Inoculated with bacteria obtained from alfalfa root
tubercles.
Pot No. 4.—Inoculated with bacteria obtained from infected sweet
clover soil.
Pot No. 5.—Inoculated with bacteria obtained from sweet clover
root tubercles.
Plate 5 clearly shows the results obtained and certainly furnishes
conclusive proof that the same effect is produced upon the growth
of the alfalfa whether the nitrogen-gathering bacteria used for the
inoculation are obtained from alfalfa soil, from alfalfa tubercles, from
sweet clover soil, or from sweet clover tubercles. It also illustrates
the importance of bacteria in growing alfalfa as will be seen by
comparing the four inoculated pots with the uninoculated pot, which
is No. 1, on the left in each series of views. The upper view was
taken when the alfalfa plants were five weeks old; the next series
when they were six weeks old; the next, seven weeks old; and the
lower series when they were eight weeks old, from the time of
seeding.
A duplicate series of pots prepared in exactly the same manner
gave similar results.
The infected alfalfa soil was obtained from a field of three-year-old
alfalfa, which was inoculated when first seeded, with infected alfalfa
soil obtained from an old alfalfa field in Kansas. About one pound of
this soil was shaken in a quart of water, the soil allowed to settle,
and some of the nearly clear solution used for the inoculation of Pot
No. 2. The alfalfa tubercles from which bacteria were obtained were
carefully washed in distilled water to free them from adhering soil
particles, and then rubbed up in distilled water, a small amount of
this water being then used for the inoculation of Pot No. 3. The
infected sweet clover soil was obtained from a place by the roadside
where sweet clover was growing luxuriantly and well provided with
root tubercles. This place was about two miles from the nearest field
ever seeded to alfalfa, so far as known. A water extract from this soil
was used to inoculate Pot No. 4. The bacteria from sweet clover
tubercles were obtained in the same manner as those from alfalfa
tubercles, and were used to inoculate Pot No. 5.
 Plate 5. Alfalfa: Effect of Bacteria from Alfalfa and from
Sweet Clover.

Pot 1.—No bacteria.

Pots 2 and 3.—Bacteria from alfalfa.

Pots 4 and 5.—Bacteria from sweet clover.

The four series of photographs were taken five, six,

seven and eight weeks from time of planting,
respectively.

From these investigations we thus have conclusive evidence that

infected sweet clover soil can be used for the inoculation of alfalfa
fields, the bacteria of the two plants acting the same. The infected
soil may be obtained from any place where the sweet clover is found
growing with abundance of tubercles on its roots. The soil may be
collected to a depth of three or four inches and scattered over the
alfalfa field at the rate of 100 pounds or more to the acre. It is well
to scatter the infected soil at about the time the alfalfa is seeded,
and harrow it in with the alfalfa seed, although it may be applied
some days or even some weeks before seeding time, and probably it
would be all right to apply the infected soil the fall before, for it is
known that the bacteria will live in soil for several months, even
though the soil be placed in sacks and allowed to become quite dry.
Investigations have shown that 100 pounds of thoroughly infected
soil to the acre is sufficient to produce a very satisfactory inoculation
within one year from the time it is applied. Of course, somewhat
heavier applications may well be made if it can be done at small
expense. The infected soil need not be applied with any high degree
of uniformity, but special care should be taken that the higher places
and watersheds are not missed in scattering it over the field. If a
few square yards, or even square rods, should be missed on the
slopes or lower land, it would make but little difference, as the
bacteria will be washed over such places from the higher land.
After the soil becomes somewhat dry it is easily scattered by hand
from the wagon or from a sack which one can carry. Sometimes it is
applied by means of an end gate seeder or a fertilizer drill, or it
could be spread by a manure spreader with an application of
manure.
The question naturally arises whether there is not danger of
getting some sweet clover seed with the infected sweet clover soil,
and thus of getting sweet clover mixed with the alfalfa in the field.
In the writer’s opinion there is little or nothing to fear in this
matter. In the first place, the amount of sweet clover seed thus
obtained would be very small, probably none at all, if one were
careful to scrape off the vegetable matter, and perhaps a half inch of
earth before collecting the infected soil (most of the bacteria are
probably between one-half inch and six inches in depth, as most of
the tubercles develop and decay between those depths); second; it
is doubtful if a small amount of sweet clover hay would lessen the
value of alfalfa hay in the least, for stock frequently eat small
amounts of sweet clover of their own choice even when it is nearly
mature, and if it is cut while still quite immature and tender it makes
quite satisfactory hay, so much so that in some sections of the
United States, particularly in the South, sweet clover is regularly
seeded on fields and cut for hay, and it is found to be a valuable and
very nutritious feed, the live stock eating it in large quantities, and
with apparent relish, after they have acquired a taste for it; third,
sweet clover is not known as a bad weed in the fields or meadows,
even where it has been a common roadside plant for many years,
and, being naturally a biennial plant, if it were cut down every five or
six weeks, as we commonly cut alfalfa during the season, it would
almost certainly die out after a few years while alfalfa, a perennial
plant, would continue to live.
Only one instance has come to the writer’s attention where alfalfa
has been growing for several years with sweet clover growing in the
field or fence rows beside it. This is on the farm of Mr. D. S. Mayhew,
of Mercer County, Illinois, who writes as follows regarding the
matter:

“Will say that the sweet clover has made no

headway in my meadow, as it did not go to seed, on
account of my cutting it so often. The sweet clover got
into the alfalfa in the seed when I sowed it. I do not
think it will do any harm in a meadow, but I believe it
would do harm in a pasture if it wasn’t cut down as
stock will not eat the sweet clover.”

Of course if sweet clover should get into the field and persist in
growing, and if it were found to injure the alfalfa appreciably or
markedly, we can always resort to plowing the ground up and
growing corn or other crops, thus obtaining some benefit from the
leguminous crop for its fertilizing value, and at the same time
completely eradicating the sweet clover, but leaving the soil well
infected with alfalfa bacteria ready to serve in case alfalfa should be
again seeded within a few years.

Conclusions.
In general agriculture in Illinois, whether it be grain farming or
ordinary livestock farming, the growing of legumes is absolutely
essential as a part of any economic system which shall maintain the
fertility of the soil; and for the successful growing of legumes the
presence and assistance of the proper species of nitrogen-gathering
bacteria are also absolutely essential. These facts being granted, it
certainly follows that when sowing any legume on land, where the
same legume has never been grown before, or perhaps where it has
not been successfully grown within recent years, we should always
consider the matter of inoculation; and, unless there is good reason
to believe that the soil has been inoculated by the washing from
other higher lying land where these bacteria are known to be
present or by applications of manure made from that legume, or by
some other such incidental means; or unless there is evidence that
the bacteria are carried with the seed in sufficient quantity to effect
a satisfactory inoculation (as appears to be the case with the
cowpea), then we should inoculate the soil directly with the specific
bacteria required by the legume which we desire to grow.
While some Illinois soils are becoming deficient in phosphorus and
in lime, especially in the southern part of the state, and while
phosphorus[8] and ground limestone can be applied to such soils
with marked benefit and profit, especially for the growing of
legumes, there is abundant evidence that one of the dominant
causes for the failure or unsatisfactory growth of some of our most
valuable legumes, and on some soils the sole cause of failure, is the
absence of the proper nitrogen-gathering bacteria.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

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

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookultra.com