OTHER BOOKS:
BY PROFESSOR LANDAU:
Differential and Integral Caleulus
Elementary Number Theory
Grundlagen der Analysis
‘Basdbuch der Lehre von der Ver-
(eilung der Primzahlon, 2 Vols.
‘Wiefuhrung in die Elementare und
Analytisehe Theorie der Alge-
‘beaischen Zahlen und der Ideale
Omeetellung und Begrandung
‘Bisger Neuorer Ergebnisse der
‘Pysktionentheorie
Werdesngen Uber Zahlentheorie,
3 Vals,
fe Zablentheorie
11 of Zahlentheorie)
FOUNDATIONS OF ANALYSIS
THE ARITHMETIC OF
WHOLE, RATIONAL, IRRATIONAL
AND COMPLEX NUMBERS
A Supplement to Text-Books on the
Differential and Integral Calculus
BY
EDMUND LANDAU
AMS CHELSEA PUBLISHING
American Mathematical Society + Providence, Rhode IslandPrimary 26-01, O1A75,
PREFACE FOR THE STUDENT
1. Please don’t read the Preface for the Teacher.
2. I will ask of you only the ability to read English and to
think logically—no high school mathematics, and certainly no
advanced mathematics.
number, no number, two eases, alll
7 ete, are completely unambiguous
phrases. “Theorem 1,” "Theorem 2”... . “Theorem 301” (and
the like in the case of axioms, definitions, chapters, and sections)
and also 2)" (used for distinguishing cases) are simply
ing the various theorems, axioms, det
tions, chapters, sections, and cases, and are more convenient for
yurposes of reference than if I were to speak, say, of “Theorem
Light Blue,” “Theorem Dark Blue,” and so on. Up to “801,” as
‘a matter of fact, there would be difficulty whatever in introduc-
itegers. The first difficulty—overeorfie
Lee
with the mysterious series of dots after the comma (in Chapter I,
they are called natural numbers), in defining the arithmeti
operations upon these numbers, and in the proofs of the pertin¢
ling material in each of the chapters in
Chapter 1, for the natural numbers; in Chapter 2, for
tive rational numbers; in Chapter
8, for the positive (rational and irrational) numbers; in Chap-
ter 4, for the real numbers (positive, negative, and zero) ; and
in Chapter 5, for the complex numbers; thus, I speak only of
such numbers as you have already dealt with in high school.
In this connection:
8. Please forget what you have learned in school; you haven’t
learned it,Paorace For rue SrUDENT
Please keep in mind everywhere the corresponding portions
of your school work; you haven't actually forgotten them.
4. ‘The multip! table is not to be found in this book,
not even the theorem
224;
but I would recommend, as an exercise in connection with Chap-
ter 1, §4, that you make the following definiti
2—141,
4= (+1) +) +1),
and then prove the theorem.
5. Forgive me for “theeing” and “thouing” you.* One rea-
son for my doing so is that this book is written partly in usum
delphinarum:} for, as is well known (cf. B. Landau Vorlesungen
liber Zahlentheorie, Vol. I, p. V), my daughters have been study-
ing (Chemistry) at the University for several semesters already
and think that they have learned the differential and integral
calculus in College; and yet they still do
nyayn.
Berlin, December 28, 1929.
Edmund Landau
Professor Landau uses the fail
[Prans.)
1 Delphin use. The Delphin classics were prepared by great French
scholars for the use of the Dauphin of France, ton of King Louis X1V.
[Prone
‘du (thou)
PREFACE FOR THE TEACHER
‘This little book is a concession to those of my colleagues (u»
fortunately in the majority) who do not share my point of vie
‘While a rigorous and complete exposition of elementary math:
maties can not, of course, be expected in the high schools, th
mathematical courses in colleges and universities should acquai
‘the student not only with the subject matter and results of math
matics, but also with its methods of proof. Even one who studir
mathematics mainly for its applications to physics and to oth
seiences, and who must therefore often discover auxiliary math
matieal theorems for himself, can not continue to take ste}
securely along the path he has chosen unless he has learned ho
to walk—that is, unless he is able to distinguish between true ar
false, betwoen supposition and proof (or, as some say so nicel
between non-rigorous and rigorous proof).
I therefore think it right—as do some of my teachers and cc
leagues, some authors whose writings I have found of help, ar
most of my students—that even in his first semester the stude
should learn what the basic facts are, accepted as axioms, fro
which mathematical analysis is developed, and how one can pr
ceed with this development. As is well known, these axioms c
be seleeted in various ways; so that I do not declare it to be i
correct, but only to be almost diametrically opposite to my poi
of view, if one postulates as axioms for real numbers many of t
ie main theorem of this bo:
T do not, to be sure, pro
the consistency of the five Peano axioms (because that can not
done), but each of them is obviously independent of the precedi:
ones, On the other hand, were we to adopt a large number of axion
as mentioned above, the question would immediately occur to t
student whether some of them could not be proved (a shrewd 0
would add: or disproved) by means of the rest of them. Since it h
been known for many decades that all these additional axioms ¢
(Theorem 205, Dedekind’s TheoretTABLE OF CONTENTS
Preface for the Student..........0. 7 v
Preface for the Scholar. .......2.c0+++ Sees ivel
Cuarren
NATURAL NUMBERS
§1. Axioms : 1
§2. Addition 8
§3. Ordering .... 9
§4, Multiplication 1
Cuaprer U1
FRACTIONS
§1. Definition and Bauivalence.......0+ceceeeeseveeeees 19
§2. Ordering a tan
E 26
a i 31
Rational Numbers and Integers. 35
Cuarren 1
curs
§1. Definition . poate se tetas 43
§2. Ordering 7 ne Saag
§3. Addition .. 48
§4. Multiplication ...... a 54
§5. Rational Cuts and Integral Cuts . 61
Cuseren TV
REAL NUMBERS
§1. Definition 22... ses 69
§2. enter 70
§3. 7
§4. 84
§5. Dedekind’s Fundamental Theorem. ....‘Tanis oF Conmenra
Cuarren V
COMPLEX NUMBERS
2
98
- 96
|. Subtraction... Hee : 101
Division : 102
mplex Conjugates 108
/- Absolute Value . sev -108
;. Sums and Products as m2
. Powers .. 126
Incorporation of the Real Numbers into the System of
‘Complex Numbers at
CHAPTER 1
NATURAL NUMBERS
gl
Axioms
‘We assume the following to be given:
A.set (i.e totality) of objects called natural numbers, possessing
the properties—called axioms—to be listed below.
Before formulating the axioms we make some remarks about
the symbols — and + which will be used.
Unless otherwise specified, small italie letters will stand for
natural numbers throughout this book.
If z is given and y is given, then
either 2 and y are the same number; this may be written
wey
(& to be read “equal
or z and y are not the same number; this may be written
ey
(+ to be read “is not ogual to”)
‘Accordingly, the following are true on purely logical grounds
» ame
for every 2.
2) It
poy
then
3) If
then2 1, Naren Nua [Axioms 15,
‘Thus a statement such as
asbac=d,
which on the face of it means merely that
a=d, b=ee=d,
contains the additional information that, say,
swing properties:
Axiom 1: 1 is @ natural number.
‘That Is, our set is not empty; it contains an object called 1
‘Aviom 31 We always hove
eK
‘That is, there exists no number whose successor is 1.
‘Axiom 4: If
var
then
z=4.
That is, for any given number there exists either no number or
‘oxactly one number whose successor is the given number.
Axiom 5 (Axiom of Induction) : Let there be given a set 3
Ie then so does 2’
Thon 1 contains all the natural numbers.
mis] 52 Ammon 8
g2
Addition
‘Theorem 1s If
eee
then
vey.
Proof: Otherwise, we would have
vor
and hence, by Axiom 4,
sey.
‘Theorem 21 we.
Proof: Let 3 be the set of all x for which this holds true,
1) By Axiom 1 and Axiom
igs to MR, then
vhs,
and hence by Theorem 1,
(yee,
so that 2’ belongs to 9.
By Axiom 5, 9 therefore contains all the natural numbers, ic.
we have for each 2 that,
vps
‘Theorem 3: If
ea
then there exists one (hence, by Axiom 4, exactly one) u such that
Proof: Let 38 be the set 1g of the number 1 and of all
for whieh there exists such a x. (For any such , we have
nocessity that
ca
by Axiom 8.)
1) 1 belongs to 5.4 1. Navona, Noses ims,
M1) If = belongs to m2, then, with w denoting the number 2,
we have
vow
so that 2" belongs to 2.
By Axiom 5, 90 therefore contains all the natural numbers;
thus for each
zel
there existe au such that
‘Theorem 4, and at the same time Definition 1: To every pair
of numbers 2, v, we may assign in exactly one way a natural num-
er, called 2 + y C+ to be read “ plus”), suck that
For every 2,
2) 24 of = (e+) for every # and every ¥.
2+ y ie called the sum of 2 and y, or the number obtained by
addition of y to 2.
|A) First we will show that for each fixed z there is
of defining x + y for ally in such a way that
rtl=z
eiv=G@ty
Let a, and b, be defined for al
a=,
for every y.
and be such that
ay = (, for every y.
Let 97 be the sot of all
9 gare
1 belongs to 1.
If y belongs to 9, then
h
4, = Py
hence by Axiom 2,
therefore
* =¢
0 that yf belongs to 2.
Hence mis the set of all natural numbers; Le. for every y we
have
4, = by
Def.) $2 Aum 5
B) Now we will show that for each = it is actually possible to
define + y for all y in such a way that
atias
and
aty—(ety) — for every
Let m he the set of all x for which this is possible (in exactly
fone way, by A)).
1) For
zed
the number
sty=r
is as required, since
atleal=z,
ety=
Hence 1 belongs to
ID) Let z belong to 9, so that there exists an z+ y for all y.
‘Then the number
+m
vtyaletyy
is the required number for 2’, since
#41 = e+)! =
and
2 ty = ebyy om (een = ety"
Hence 2" belongs to 8.
‘Therefore ® contains all =
‘Theorem § (Associative Law of Addition) :
Gtyteartyte.
Proof: Fix 2 and y, and denote by 98 the set of all for which
the assertion of the theorem holds,
(tM tl=Gt erty eat FD:
thus 1 belongs to am.
I1) Let = belong to 3. Then
(etyte=24ta,
hence:
(ayes = (ety) tay = (outa = 4 tey = ete n
0 that 2 belongs to
‘Therefore the assertion holds for all z.6 1 Nasuna, Nuwpens Imo]
Theorem 6 (Commutative Law of Addition) :
ehusyta
Proof: Fix y, and let ® be the set of all x for which the
assertion holds
1) We have
veiled,
‘and furthermore, by the construction in the proof of Theorem 4,
lty=v,
0 that
Ltyeytt
and 1 belongs tom.
M1) If x belongs to 28, then
stysyts,
therefore
(tate ete.
By the construction in the proof of Theorem 4, we have
vt y=ty
hence
etymvew,
0 that 2° belongs to 9.
‘The assertion therefore holds for all z.
‘Theorem 7: vhoty
Proof: Fix x, and let 3 be the sot of all y for which the asser-
tion holds.
nD L4e,
Lee td;
1 belongs to %.
If y belongs to wi, then
wee
hence
0 that y’ belongs to a.
‘Therefore the assertion holds for all y,
Theorem 8 If
then
- etybete,
$2. Avorniox 7
Proof: Consider a fixed y and a fixed such that
us
fand let SP be the set of all x for which
rtypete,
v wee,
Ltubites
1 belongs to m2.
If x belongs to 3, then
rhypets,
hence
(+e to,
Bybee te
so that 2" belongs to Wt,
‘Thorefore the assertion holds always,
‘Theorem 9: For given x and y, exactly one of the following
must be the case:
> ae.
2) There exists a u (exactly one, by Theorem 8) such that
roy tu
8) There exists a v (exactly one, by Theorem 8) such that
yurte.
A) By Theorem 7, cases 1) and 2) are incompat
peytun(etetumrt(osm—wtwte
‘Therefore wo'can have at most one of the cases 1), 2) and 8)
B) Let a be fixed, and let 5 be the set of all y for one
(hence by A), exactly one) of the eases 1), 2) and 8) obtains,
1) For y=1, we have by Theorem 3 that ei
aoiny (case
eew=l
Hence 1 belongs to sx.
II) Let y belong tom. Then
‘either (case 1) for 7)
=ute (case 2))
z=8 1. Navona, Nusa [re 20.12,
hence
Yoytlaztl (case 3) for v5
or (case 2) for y)
eau
hence it
u=1,
‘thon
zeytimy (ease 1) for 7);
but if
wel,
then, by Theorem 3,
usw=lte,
ray t (tw) a(ytl) twa tw
(case 2) for u's
or (case 3) for ¥)
verte,
hence
valet mete
(case 8) for v7).
In any case, yf belongs to
‘Therefore we always have ono of the eases 1), 2) and 3).
Def, 25} $8. Onmaine °
83
Ordering
Definition 2: If
zeytu
then
z>u
(> to be read “is greater than.”)
sf
(< to be read “is less thar
‘Theorem 10: For any given x, y, we have exactly one of the
onsee
=n tH FK4
Proof: Theorem 9, Definition 2 and Definition 8.
Theorem 1: If
m>y
then
vce,
Proof: Each of these means that
zoytw
for some suitable «
‘Theorem 12: If
acy
then
ype.
Proof: Each of these means that
yarte
xem
e>y or ry
( to be read “is greater than or equal to.”)
Definition 5: rsy10 1. Narunst, Nuamens (mas)
acy or tmy.
(S to be read “is less than or equal to.")
‘Theorem 13: If
then
Proof: ‘Theorem 11.
Theorem 14: If
then
Proof: Theorem 12.
‘Theorem 15 (Transitivity of Ordering): If
By U>H
then
Be
ren 9a,
Proofs etyaety
Theorem 19: If
2>y or eayoracy
then
ste>ytsorstimyts or ztrcy ty
respectively,
Proof: 1) It
then Ae
saute
rhe my tete = ety) te = uty td) = Otel tH,
thesyts
2
aay
then clearly
chem yte
3) If
then aces
y>s,
henee, by 1),
peers,
zhecyte
Theorem 20: If
aHEDyts oratemyty oretscyts,2 1, Navona, Nestuens (ther
cis 2>y remy ore
Proof: Follows from Theorem 1
both instances, mutually exclusive and exhaust all possi
Theorem 21: If
sinee the three e
r>y 2>u
ate>ytu
Proof: By Theorem 19, we have
ate>yte
and
vtemety>utuaatu,
etepyte
2 If
5Eyr>uorr>y2By
phe>y tm
Follows from Theorem 19 if an equality sign holds in
sis, otherwise from Theorem 21
yrem 23: If
Proofs 0
us if two equality signs hold in the hypothesis;
orem 22 does it.
‘Theorem 24 rz
Proofs Either
naw=uti>l.
Theorem 251 If
woe
yert1.
Proof: vertu
«zh
hence
$9. Onmause 12
vert.
‘Theorem 26: If
vert
then
vse
Proof: Otherwise we would have
w>e
and therefore, by ‘Theorem 25,
vert.
rery nonempty set. of natural numbers
one which is less than any other number
‘Theorem 27: In
there ie a least one
of the set).
Proof: Let ® be the given set, and let 1 be the set of all 2
are = every number of 3
rs to MR; in fact, for each y of K the number y +1 does
not belong to SR, since
wtlow
‘Therefore there is an m in ® such that m +1 does not belong
to m; for otherwise, every natural number would have to belong
to Wi, by Axiom 5.
Of this m I now assert that tia < every nof R, and that it belongs
to. The former we already know. The
indirect argument, as fol
each n of Swe would have
men
hence, by Theorem 25,
m+isn;
thus m + 1 would belong to mt, contradicting the statement above
by which m was introduced.2 IE Faacntons
since
‘Theorem 114: If Z is the rational number corresponding to
Definition 271 The U of Theorem 110 i ealled the quotient of
X by Y, or the rational number obtained from division of X by Y.
1 wi be ented by & (tbe read oer
Let X and ¥ be integers, say X =x and ¥—y. Then by Theo-
rem 114, the rational number & determined by Definitions 26 and
27 stands for the class to which the fraction 5 Gin the earlier
sense) belongs.
We need not be afraid of confusing the two symbols Q since
fractions as such will from now on no longer occur. % wi
y
forth always denote a rational number. Conversely, every rational
henee-
‘number may be expressed in the form 5 » by Theorem 114 and
Defition 2.
Theorem 115: Let X and ¥ be given Then there exits a 2
wah that
y aX > ¥.
roots Z- iva rafonal number; by Theorem $9, there exist
Integra (i our nw terminology), say # and, seh hat
es
ork
By Theorem 111, we have
enc, by Theorem 1065,
Kea k= x(S0)= (x4 oz(xe)a—xtoxd =v,
Def, 27-291 $1. Depron «
CHAPTER It
curs
gl
Definition
Definition 28: A set of rational numbers is called a cut if
1) it contains a rational number, but does not contain alt
rational aumbers;
2) every rational number of the vet i smaller than every
rational number not belonging to the set;
does not contain a greatest rational number (i.e. a number
greater than any other number of the set)
Iso use the term “lower class” for such a set, and the
I rational numbers which are
jwer numbers” and “upper numbers,” respectively.
‘Small Greek letters will be used throughout to denote cuts, except
where otherwise specified
Definition 29: py
(<= to be rend “is equal to") if every lower number for § is a lower
number for and every lower number for » is a lower number
for &
In other words, if the sots are identical.
Otherwise,
:
(Cf to be read “is not equal to
‘The fol wree theorems are trivial:
Theorem 116: fae
‘Theorem 117: If fa
then
a8
Theorem 118: If f= 4, 9=%,
then
gmt