Professional Documents
Culture Documents
llli
:
;
V-iii;
s;;;
piill
Wm
PRESENTED BY
Vrof, -H ovsf o-rd_
FAMOUS PKOBLEMS
ELEMENTARY GEOMETRY
THE DUPLICATION OF THE CUBE
THE TRISECTION OF AN ANGLE
THE QUADRATURE OF THE CIRCLE
im
and
in
/
Boston, U.S.A., and London
JJre
Copyright,
1897,
by
K6*
PREFACE.
The more
precise definitions
of
of
gymnasium
me
With a view to
counteract-
summer
than
Some time
before, I
and
an
summer
semester.
my
been
I ven-
lectures to
Tgert, of Ems,
tioned.
who attended
out under
students.
tribute to
may
con-
KLEIN.
TRANSLATORS' PREFACE.
At
German Association
for
the duplication
modern
circle, as
research.
work
of the
gymnasium.
is
likely
tions into
The treatment
of the subject
knowledge of the
required.
Among
differential
is
is
scendental numbers?
How
it
be effectedTvVhat^are tran-
ir
are
transcendental ?
With the belief that an English presentation of so important a work would appeal to many unable to read the original,
TRANSLATOR'S PREFACE.
vi
In
its
made
free use of
following
They
its
modifications where
desire further to
it
seemed advisable.
assist-
W. W. BEMAN.
August, 1897.
D. E. SMITH.
CONTENTS.
INTRODUCTION.
PAGE
PART
The Possibility
Chapter
7, 8.
9.
10.
11, 12.
13, 14.
I.
Normal form
of x
be constructed
f (x)
F(x)
....
Conjugate values
...
Chapter
...
.
I.
10
of 2
.11
1.
2.
3.
The
Chapter
1.
III.
2-4. Gauss's
prime numbers
13
straight edge
14
13
into
Equal Parts.
.......
16
17
5.
19
6.
Gauss's
Lemma
19
7, 8.
The
.21
CONTEXTS.
viii
Chapter IV.
1.
2-4.
5, 6.
7.
24
25
compasses
8, 9.
Von
27
and
32
Chapter V.
34
1.
Paper folding
42
2.
42
4.
5.
Mechanical devices
3.
44
45
.
PART
.47
II.
Chapter
I.
Demonstration of the
Transcendental Numbers.
Cantor's
1.
Definition of algebraic
2.
Arrangement
3.
Demonstration of
Chapter
1.
2.
3.
4, 5.
II.
49
50
53
to 1770
III.
of
of algebraic
Chapter
59
1.
2.
3.
Existence
<p(x)
e.
.... .61
.
62
65
CONTENTS.
Chapter IV.
ix
it.
PAGE
1.
68
2.
The function
70
3.
Lindemann's Theorem
Lindemann's Corollary
4.
5.
6.
7.
^(x)
The transcendence
The transcendence
The transcendence
Chapter V.
of
73
74
:
ir
=
of y =
of y
76
ex
77
sin 2 x
77
1.
The
2.
3.
Geometric construction of
7T.
78
.
-n-
78
79
INTRODUCTION.
is
my
part to
-Still it is
not
On the
other hand,
it
presupposes but
little
preliminary work,
as, for
example,
We
case
as
to
much .to
solution.
The problem of
They
are
In
ir.
is
INTRODUCTION.
The impossibility
is
was
and second
That of the
first
On
we
find
practical
insist
upon the
constructions.
difference between
For example, if we
(i.e.,
other numbers.
is
purely
No
practical advantage
is
theoretical.
The same
is
constructions are,
each case.
1.
2.
3.
4.
We
shall consider
and compasses.
INTRODUCTION.
In analysis we have
first
rational
and
operations
addition,
These operations
can be directly effected geometrically upon two given segments by the aid of proportions, if, in the case of multiplication and division, we introduce an auxiliary unit-segment.
subtraction, multiplication,
division.
and transcendental.
The simplest
algebraic opera-
On
is
In
therefore
an analytic
constructed with straight edge and compasses
sufficient condition that
expresis
that
it
we wish
if
to
INTRODUCTION.
fortiori the
solution
is
impossible
when
the problem
An
expression
which satisfies no algebraic equation is called a transcendenThis case occurs, as we shall show, with the
tal number.
number it.
PART
I.
CHAPTER
I.
of alge-
If
x,
expressions
tion f (x)
1.
To
suppose
2.
0,
is
x,
of the form
= Va + VcTeT-f-Vd + Vb
Va + Vb
p_+Vq
?
Vr
a, b, c, d, e,
f,
The number
any term of
p, q,
upon rational
roots, it is
whose degree
it, e.g.,
where
and square
Let
fx
x is called
designate the
//,
maximum
the preceding
order, so that
no term
FAMOUS PROBLEMS.
In the example
4.
expressions of the
x
it
really depends
= V2 +
first order,
V3 + V, we have
it may be written
terms of
so that
x,
three
but as
among
made
fx
in all the
none can be
/x
or of lower order.
We
make
shall
the order
fx
This hypothesis
plicitly or implicitly.
is
obviously a very
Normal Form of
5.
x.
sum
we may reduce them
If the expression x is a
inators
to the
Suppose
VQ
VQ
and Q, which
is
fx
is
VQ may
the
x of
order
fx
maximum
it
can occur
Since,
order.
be expressed as functions of
we may put
= a + b VQ
X_
c+-d VQ'
where
a, b, c,
d contain no
more than
terms of order
/a,
we may
VQ,
VQ
write
=a +
:
/?!
VQi,
etc.
'
/x
we
tt.
ALGEBRAIC EQUATIONS.
The
may,
We
may
its
a term
linearly.
occur, for a
We
ft.
aud
still
depend upon
1 terms
ft
of order
= au + a
VQi,
12
= u + u VQi,
and hence
x
6.
We
= (au + a
12
VQi)
proceed in a similar
1,
which occur
way with
and in Q, Q 1} etc., so
that each of these quantities becomes an integral linear function of the term of order p
1 under consideration.
We
then pass on to terms of lower order and finally obtain x, or
rather its terms of different orders, under the form of rational
of order
fx
explicitly
We
Let
is
reduced to
we
all possible
ways,
we have
thus, if
= ^a + Vb + ^a Vb,
Vb.
all distinct
FAMOUS PROBLEMS.
If x is
9.
nomial
F
F
(x)
is
= (x x
(x)
clearly
jugate values.
It
x)
(x
2)
(x
x 2m
),
roots (8).
The
to x
coefficients
of the polynomial F
(x)
are rational.
For let us change the sign of one of the square roots this
permute two roots, say x A and x A since the roots of
;
will
(x)
-,
As
we merely change
the polynomial
is
A ) (x
these
Xy),
Hence
(x).
not changed.
10.
When any
coefficients, f (x)
0, the
same
is
true of
f (x) is
(x),
= +
Let x x
a
VQ be one of the conjugate values VQ, a
term of order fx a and now depend only upon other terms
of order ll and terms of lower order.
There must, then, be a
;
conjugate value
Xl
Let us
into the
'
= a -y3VQ.
=
= A+BVQ;
f (x x )
may
be put
ALGEBRAIC EQUATIONS.
when A and B
are simul-
taneously zero.
VQ = -|r
i.e.,
VQ
of order
which
all
is
//,
B,
= A-BVQ;
so also (x/) = Whence the following
f( Xl ')
hence
if f (x x )
proposition
If
0,
0.
x x satisfies the
The proof
f (x)
from
xx
equation
0, the
/x.
obtained in an
is
may be done
without affecting the generality of the reasoning, that the
analogous manner.
VQ'.
f (x x )
(a)
If x x
may
( Xl )
fx,
VQ
and
= p -f q VQ +
VQ'
+ VQ VQ7 = 0.
s
p,
we should
only have to add to the preceding expression a greater number of terms of analogous structure.
Equation
(a) is possible
(b)
Otherwise
VQ
0,
only
0,
when we have
r
0,
separately
0.
FAMOUS PROBLEMS.
10
VR
and VR'
and
quantities,
if,
VR
for the
and VR',
we have
(c)
= + A VR +
,
fx,
+ v V R V R' = 0,
VR'
0,
0,
ft
0,
0.
VR
ft
ft
and VR'.
f
(x)=0
from
is
1.
1 1
We
(x)
and
f (x)
0.
equation,
<
(x)
0,
12.
I.
<f>(x)=0
is
<
an irreducible equation,
(x)
i.e.,
<j>
(x)
For
if
we had
*(x)=*(x)x(*),
=
x^s.
Xj
=
<j>
and conse-
0.
(x) cannot be
This irreduci-
is
the rational
ALGEBRAIC EQUATIONS.
then
(x x )
<f>
But
or both..
11
= 0, or %
= 0,
x
( i)
by
all
the
<f>
(x)
has no multiple
Otherwise
roots.
<f>
(x)
could
(x)
<f>
and
(x)
would not be
We
Let
IV.
and
and
<f>
(x)
Otherwise
x^s.
divisor,
which
would not be
irreducible.
let
We
be these quantities.
<t>
For $
(x)
= C (x xj (x
(x)
shall then
is
satisfied
have
x 2)
(x
xM
by the quantities
xA
).
and
it
has no
multiple roots.
for
irreducible.
(x)
<f>
(x)
<f>
Y.
<f>
<f>
<f>
(x)
is
coefficients satisfied
by the
x/s.
For
if f (x)
by
by
By
divisible
lished,
we may
<j>
x
<f>
were another
and consequently
(x) and therefore
(x)
thus estab-
(x)
and
<f>
Xj's
multiple roots.
(x).
(x)
= F, (x)
cf>
(x).
by
cf>
(x)
that
is,
FAMOUS PROBLEMS.
12
by dividing F
tient obtained
constant
one
by
it
<f>
admits
all
<f>
Fx
(x)
If F x (x)
(x).
(x)
Hence F x
and
(x),
by
(x)
it
=F
(x)
not a
is
and admitting
(x).
and
_!
F(")
minimum
14.
is
degree
We
(x),
is
also a
(x),
is
it is
2m
Therefore
<f>
further,
theorem
The degree of the
= C- [()]'.
now determine
can
of degree 2 m
<ft
= Ci
(x)
for F (x),
the degree
of <(x).
<f>
(x).
(x)
Hence
= v M.
15.
Since,
is
equation satisfied by
all
an expres-
always a poiver of
is
2.
we have
the converse
theorem
If an irreducible equation
solved by square roots.
is
not of degree 2 h ,
it
cannot be
CHAPTER
II.
of the Angle.
1.
The equation
of the cube.
of the
problem
is
manifestly
x
2.
3,
v2 would
have a
For an equation of the third degree which is
reducible must have a rational linear factor.
Further, the
degree of the equation is not of the form 2 h hence it cannot
be solved by means of square roots, and the geometric construction with straight edge and compasses is impossible.
This
irreducible,
is
since
otherwise
rational value.
2.
Xext
let
x
A.
A,
of the form a
may
be a complex quantity
ib.
The question
arises
whether one of
tion of
It
X.
its
an expression always depends upon the values of the quansupposed to be known. In the case x 3
2, we were
dealing with numerical quantities, and the question was
tities
3,
whether
V2
equation
a rational function of
A..
In the
first
In the
represented by
case,
the so-called
FAMOUS PROBLEMS.
14
which
in
<j>
equation.
whence
~"
(cos
<f>
lm
We
as arbitrary.
The
cfi
sin
<,
<f>
can-
gin
<f>)
sin
into
itself
<f>.
two,
to
number and
Vr,
it
representing by
>
Assume
A.
cos
ty\=tyr Vcos
also that of a
FlG
I.
\<P
therefore
is
3.
satisfy our
is
and
- 1 +2
*=-
Vr,
are
Vr,
V3
e
1
= - 1 -2
V3
*
The
are,
= cos +
by De Moivre's formula,
<
sin
<
= cos -+
x2
= cos^
f-ism^
= cos -
x3
<
sin-,
|-
15
sin
a_cJL4
jj
J!*
The
shows that
Hence
-^.
the equation
x
is
= cos +
<j>
si
<
2.
function of cos
on substituting
<f>
and sin
<j>,
for
by a continuous variation
its
<.
But
if
we
of the angle
<f>,
effect this
we
change
x3
<f>.
cial values of
e.g.,
when
<
<
= .
the construction
may
prove to be possible,
CHAPTER
The Division
III.
Equal Parts.
parts has
Gauss extended
Arithmeticae,
Disquisitiones
we make
known to
/*
and
1,
the ancients.
-J-
In his
of
series
If in p = 2 +
=
we get p
3 and 5, cases already
For = 2we get p = 2 + 1 = 17,
=2
is
this
their powers.
22
/i
For
fx
The
ber.
Similarly for p
fi
5,
fi
6,
/a
The proof
is
fi
we have
numbers corresponding
to
//.
5, 6,
It
number
is,
fx
=4
is
the
last
for
memoir
is
De
257
resolutione algebraica aequationis x
1, sive
17
of
We may
restrict the
= a/ + bv;
a
b
1 =--f--.
whence
[IV
To divide the
know how
Thus, for
circle into
to divide
n
\iv
into
it
= 15, we have
[X
= n equal
p and
parts
it is
sufficient to
JL_2_3
15~~ 3
3.
As
5'
a prime number)
p
= 2h +
We
1.
shall next
when
Fermat's Theorem
divisible by p,
these
ap
*p 1
is
=+l
(mod.
p).
given value of
exponent
-1
it
particular, if
a, satisfies
may
s
the congruence.
=p1
we
If
s is
the lowest
In
1.
a divisor of p
say that a is a primitive root of p,
be shown that
s is
FAMOUS PROBLEMS.
18
We
root.
shall
make use
is
a primitive
(1)
and
2s
(2)
From
= 2- + l,
(1)
2h
<p
=+1
from
(2) 2
shows that
h is
(2)
and
(3),
by
p).
=1
~h
(mod.
p).
s-h^h, s^2h.
.-.(4)
(3),
(mod.
division,
2s
From
=-1
2h
(3)
From
p).
p.
s>h.
'.-.
(1)
(mod.
>
by squaring,
2 211
=1
(mod.
p).
Comparing with
2X
=1
(mod.
s is
p),
we have
s^2h.
(5)
.-.
We
is
= 2h.
= 2h
true of
are
the same
Hence prime
1.
This conclusion
pose that
h is
divisible
may
be established otherwise.
then,
-f-
= h'(2n + l);
2n
+l=
+l)
(x
211
2n
-!+ ., x + 1),
.
Sup-
= 2 h(2n+1) + l
is
divisible
by 2 h
+ 1,
'
19
and hence
is
hot a
prime number.
5.
We
now
reach, our
fundamental proposition
of the form
= 2* + l = 2 +
t
2'
To
(z
.=
l.
a circle of radius
iy)
1.
1, is
-1=
=1
by dividing by
1, which
0.
let
is
the equation
z
n-l
_|_
n- 2
-f z
We
thus obtain
+ = o,
_i
equation.
As noticed
we may confine our attention to the cases where n is
a prime number or a power of a prime number.
We shall
first investigate the case when n = p.
The essential point of
the proof is to show that the above equation is irreducible.
For since, as we have seen, irreducible equations can only be
solved by means of square roots in finite number when their
above,
degree
is
possible
a power of
when
Thus we
see
why
2,
is
=jfc
2h
+ 1 = 2^ +
is
always im-
2, i.e.,
when
1.
exceptional position.
6.
At
Lemma.
this point
we
introduce a
lemma known
as Gauss's
If
F(z)
= z m + Az m - +
:
Bz m
-2
+.
+ Lz+
M,
FAMOUS PROBLEMS.
20
where A, B,
are integers, and F(z) can be resolved into
two rational factors f (z) and (z), so that
.
cf>
(z)
= (z- + a
1
f (z)
m
(z "
(z)
/3 1
"- 1
zm
lZ
m -1
+ a z m '- +
2
.)
82 z-"- +...),
then mnst
a b F(z)
fx (z)
(b
The
= (a z + a^'" +
ra
fa (z)
'
.)
since a
divisor of a b
and b k the
first coefficient
of f x (z)
by q. Let
and consider the coefficient
(z)
<f> 1
a i"k
a i _ 1 b k+1
-f-
a i _ 2 b k+2
all
-|-
first
are
absurdity, since
we have
seen that
all
to
one another.
from
21
and
and b
/3's
are in-
is irre-
different
1,
a's
tegral.
7.
ducible,
first
is
it
show by Gauss's
sufficient to
Lemma
that the
coefficients.
=x+
We
XXXIX,
p.
167, which
l.
obtain
by
An
is
always p
first
itself,
by
is
ilways irreducible.
For
if this
f (x
+ 1) = (x m + a^-(x m '+b
m '-
+....+
+...+b
a m _! x
m '_ 1 x
+ bm
+ aJ
.),
we have a m b m
ors of a m b m
'
p.
must be unity.
a ra
= p,
of the fac-
Suppose, then,
bm
= l.
PiP!Zi)_ a m-l
O
we have
x,
b
D m'-lD m ,_i_ a m b
,
FAMOUS PROBLEMS.
22
The
first
divisible
member and
by
p,
a m _ 1 bm
must be so
Equating the
Since b m
also.
'
= l,
by p.
coefficients of the terms
2
in x we may show that a m _ 2 is divisible by p.
Similarly
we show that all of the remaining coefficients of the factor
111-1
m
x
But
a m _! x -f- a m are divisible by p.
axX
this cannot be true of the coefficient of x m which is 1.
The assumed equality is impossible and hence the cycloa m _! is divisible
tomic equation
8.
We now
is
is
sible.
when
irreducible
=p
We
a
.
a prime.
n
is
a power of a
the circle into p 2 equal parts is imposwill then be solved, since the
It
is
now
7^T =
zP
the roots of
z p(p-2)
we
obtain
-l
may
This
i.e.,
z p(p-d
parts,
-.
be written
+ zp 4-
i.
=o-.
=x+
l,
we have
(x
+ 1)p(p-D + (x + 1)p(p-2) +
+ (x + 1)p + 1 = 0.
23
where ^
(x) is
constant term
sion
is
We
is 1.
always irreducible.
equation
is also irreducible.
The same
equal parts
is true,
when
>
2.
CHAPTER
The Construction
We
1.
IV.
of the Regular
Polygon of 17 Sides.
how
circle into
is
possible
It will
now
effected.
The purpose
elementary
gon of 17
way how
show
in an
sides.
we
Since
of
all,
16
+x +.
+x +x+l = 0,
2
15
We
know
form
e
and
= cos 2k7T +
2k7T
sin
1, 2,
2*7T
16)
if
= cos +
2-7T
ci
that
(k
eK
sin
jy,
1,
The
center.
traction
it
leparture.
K is
is
selection of
fixed
The fundamental
2.
he 16 roots
34
the following
is
number a is said to be a
when the congruence
property
= 9
= 10
= 13
3 5 EE
36
37
38
primitive root
= + 1 (mod. 17)
33
which completely
root to the
a8
3 X EE
e 1}
stated, a
the modulus 17
iesses this
as the point of
As already
some
eK .
Forming a primitive
e 1?
letermines
;o
upon
25
e 1 is
essential to indicate
Having
17 SIDES.
= 17 1 16.
for
The number 3
pos-
we have
=14
=
3U =
3 13
3 14
3 15
3^=
3 16
39
=15
= 11
= 16
10
=12
=
=
=
2
6
(mod. 17).
Notice that
if
is
= 17q +
= e/ =
3*
rhence
f
r'
is
eT
ex
we have
we have
r,
3
.
similarly
= / = (*/). =fe).
+1
er.
fence in this series of roots each root is the cube of the preceding.
FAMOUS PROBLEMS.
26
calculated successively as
The process
The
ally arranged
periods
tions.
may
p,
and the
The degree of these last depends upon the prime fac 1. They are not necessarily equations of the
tors of p
second degree.
The general
and
also
by Bachmann
by
in his
in the following
ically
Xl
x2
= 9 + 13 + 15 + 16 + 8 + 4+ 2 +
= 3 + 10 + 5 + 11 + 14 + 7 + 12 +
Operating upon
of 4 terms
Xl
and
yi
y2
y3
y4
z1
z2
z3
z4
6.
we form 4
periods
= 13 + 16+ 4+
= 9 + 15+ 8+ 2,
= 10 + ll+ 7+
= 3+ 5 + 14 + 12.
x 2 in
1,
1,
6,
way upon
= 16 +
= 13 + 4,
-15 + 2,
= 9 + 8,
l,
the
y's,
z5
z6
z7
z8
we
obtain 8 periods
= ll+
= 10+
= 5 + 12,
= 3 + 14.
6,
7,
now remains
to
show that
17 SIDES.
27
sum
4.
z is
er
>
e 17
_r
= cos
= cos
'2ir
(17
jy
= cos
2tT
sin (17
sin
r)
2iTT
9
r)
2-7T
sin
rjy+i
Hence
er
Therefore
all
+ v = 2 cos
we
readily obtain
zt
= 2O cos,
z2
= 2 cos 4 j^,
z6
= 2 cos 7 ^,
z3
= 2 cos 2 j|,
z7
= 2 cos 5 j|,
z4
= 2cos8t^,
z8
= 2 cos 3jy.
27T
Moreover, by
z5
= 2 cos
2-rr
6,
definition,
yi
= Zi + z + z + z
= zi + z y = z + z
5.
It will be necessary to
Xl
2,
= z + z + z -f z
= z + z y =Z + Z
x2
4,
into
4,
y3
6,
8,
8.
We divide
S2
S J7 the distances
FAMOUS PROBLEMS.
28
A 1? A 2
A 17 from the
S i7 being equal to the diam-
^ 4lA
A*
Hence
= 2sin 34 = 2cOS
That
may
this
17
2tt
2 cos
a,4
,
:
the values
The
figure
5~
z6
= Si,
:== Sg,
we find
Hence
1, 2, 3, 4, 5, 6, 7, 8,
Z l == Si3,
Z4
be identical with
= 17
= 17 4h.
3, 7, 11, 15.
Z2
Z3
3T^-
we must have
4h
The angle
2.
TT
.
k+p
9, 5, 1,
2/C7T
Giving to
values 13,
equal to
i.e.,
A k A 17
Z7
Z8
Si5,
=
=S
S7
for k the
Sn,
S3
5
hence the I
Moreover, the chord A K A K + p subtends p divisions of the semithe triangle OA^A^ + p shows
is equal to S
p
circumference and
that
K
+p
Ok
-+"
Op,
and a fortiori
^k + v
^k + t ~r ^p + re-
y,
we
y
yi y
yi y
y y
y y
yi
2
2
y3
Hence
we
13
o,
yi-
0,
obtain in a similar
We now
15
x2
6.
29
= S + Si S + S > 0,
= Sis + Si + S + Sn > 0,
= Sis + Si + S S >
= S Sis + S + S n > 0,
= S Sis + S S <
y* = S Sn + S S < 0.
y < y < y <
2
Finally
17 SIDES.
way
<
Xi.
= 2 cos
JjTT
propose to calculate z x
After mak-
we can
easily deduce
In order to find
we proceed
to
period
yi,
we
have,
first,
zi
zxz2
+z =
2
We
yi.
have
&K
+ p*
p,
remem-
Thus,
Hence
it
z 1 z2
Therefore z 1 and
= 12 + 3 + 14 + 5 = y
z 2 are
z
(0
whence, since
zx
>z
4.
- yi z + y = 0,
4
2,
_
yi + Vyi^ - 4y
2
r
Z
l
-f-
4
?
Z2
_
yi-
2
Vyi -4y,
FAMOUS PROBLEMS.
30
We
yi
+y =
2
xi."
Then,
yi y 2
=(13
+ 16 + 4 + l)(9 + 15 + 8 + 2).
sum
that
is,
to
1.
Xl
-V + 4
Therefore y x
y*-xiy -l = 0,
(,)
>y
_ + Vxx + 4
yi-
whence, since y x
2,
Xl
y2
'
_
~
Xl
Similarly,
y3
+y =x
and
y 3y 4
Hence
1.
W)
whence, since y 4
y4
It
_
-
x2
>y
-x y-l=0;
2
3,
+Vx + 4
2
y8
'
_
~
x2
-Vx + 4
2
2
x2
'
Since x x
+x
is
equal to the
xi
+ x = 1.
2
Further,
Xl x 2
= (13 + 16 + 4 + 1 + 9 + 15 +8 + 2)
(10 + 11 + 7 + 6 + 3 + 5 + 14 + 12).
= 4.
xx
and
x 2 are
Xl
x1
>
-fx-4 = 0;
x 2,
_ -l +o VTf
Solving equations
by a
31
whence, since
17 SIDES.
rj,
X2
rf,
_ - 1 - Vl7
o
determined
in succession, z x is
we
upon the
Vl7,
vV +
4,
Vx 2 2 +
4,
V yi -4y 4
2
Vx! 2
Vx 2 2
+ 4=y y
+4=y y
x
3,
3.
-y
- ys) =
(y 4
2)
(y 4
y*) 2
(x,
- x ),*
3
(13 + 16 + 4 + 1 - 9 - 15 - 8 - 2)
(3 + 5 + 14
+ 12-10-11-7-6)
= 16+ 1 + 10+ 8- 6- 7- 3 2
+ 2+ 4+13+11- 9-10- 6- 5
+ 7 + 9 + 1 + 16-14-15-11-10
+ 4+ 6 + 15 + 13-11-12- 8- 7
- 12 - 14 - 6 - 4 + 2 + 3+16 + 15
- 1- 3-12-10+ 8+ 9+ 5+ 4
-11-13- 5- 3+ 1+ 2 + 15 + 14
- 5 - 7 - 16 - 14 + 12 + 13 + 9 + 8
= 2(16 + l + 8 + 2 + 4+13+lo+9-10-6-7-3-ll-5-14
-12)
2(xi
x 2 ).
FAMOUS PROBLEMS.
32
that
is,
Vx x2 + 4 Vx 22 + 4
= 2 Vl7.
tained directly.
Eeplacing now
x l9
y lt y 4
obtain in succession
*i=
-1 + V17
>
y4=
The
completed.
is
now
We may
In the geometry of the ancients the straight edge and com; the difficulty lay merely in
17 SIDES.
33
lines.
Whether the several steps in
made with straight edge or with com-
On
(Halle, 1880).
Soon
after, the
straight
edge.
1774.)
We
authors mentioned.
shall say
With
Les applications de
how
is rational.
He
which he shows
cases with the
many
work
in surveying.
first,
it is
sufficient
FAMOUS PROBLEMS.
34
npon square
being given.
cele-
fhrt mittels der geraden Linie und eines festen Kreises, als
Lehrgegenstand fr hhere Unterrichtsanstalten und zum Selbstunterricht.
To construct the
8.
we
shall
XXIV,
nal,
Vol.
LXXV,
1872).
polygon of 17 sides
is
we
will
we can
is
used.*
At the extremities
we draw two
(Fig. 4)
tangents,
and
select the
axis of X,
lower as the
perpendicular to
axis of Y.
the
0.
px + q =
4.
x x and x 2
upon the axis of X.
Lay off upon the upper tangent from A
and
as
is
+ y(y-2) =
Let
Fig.
it
x2
ment measured by -
upon the
axis of X
from 0, a segment
P
*
L. Gerard
17 SIDES.
35
the line 3 and project the intersections of this line with the
from
circle
A,
by the
and
x2
lines 1
off
Proof.
we have
2x
+x
2x
+x
upon the
1,
axis of X, x x
and
x2,
(y-2)=0;
(y-2)
of the line 2,
we multiply
If
the
x2+
members
first
x^
get
0.
two equations we
of these
_ 2) + x_^ (y _ 2)2 =
x(y
formed by 1 and
we
()
Subtract-
2.
obtain
^px(y-2) + ^(y-2)*-y(y-2) =
This
is
the equation of
intersections of
y~ + ~4~
Xl
"I
X2
which
xi
H~
x2
is
the
=P
X1X2 /
(y
r>\
~~ 2
an(i x i x 2
q>
^* +
we
circle.
y 2,
this equation
ing to the
0.
From
correspond-
~ y = A>
If
3.
we now make
get
j(y-2)-y = 0,
off
from the
line y
=2
the seg-
FAMOUS PBOBLEMS.
36
ment
4
-
line y
P
correctness of the construction
9.
is
q
the segment P
Thus the
established.
now construct
They are (see
x
()
pp. 29-31)
0/)
()
x2
xx
yx
y4
zx
(v)
~~ I
y ~~ x iY
>x
>y
>y
>z
2,
2,
8,
2.
whence
it
is
= 2 cos
x i? x 2> yi?
We
it is sufficient
to construct
y<t
then lay
tangent, y
off
'
all
j4
xi
x2
yi
X,
+
may
2,
This
We
x,'
x,'
4 upon
the
The
y/
17 SIDES.
37
(0, 2),
x 1? x 2 of
the
To
we have
to lay off
above
xi
and
below.
Xl
To determine the
point
first
we connect
xx
upon the
axis of
xi
The equation
of the line
2x
from A
to x x (Fig. 5),
+ x y 2x
x
l5
circle,
x
+ y(y-2) = o,
2x t 2
4 Xl
X!
The equation
4' Xl *
+ 4'
y=2
3utting off
upon
x>
= 2 the intercept X4
l
We
)f
construction
)f
we have obviously
the lower range one, and only one, point of the upper, so
lihat
to the point x
versely.
oo
x'
= 0, and con-
FAMOUS PROBLEMS.
38
the equation
x'
x'
must
satisfy
const.
.
=2
constant =
Since
x'
when
= 2,
as
is
4.
X1
+i
o
Fig.
6.
To determine
xl
6).
we connect with
diameter
+1
upon the lower tanThe point thus determined upon the vertical
the point
the point
This line
above.
xi
cuts off
upon the
line from 4 to
X the intercept
axis of
For the
xi
-f-
1,
5y
+ 2x = 2,
Hence
to
this point is
xi
5y
and
its
- 2x x = 2,
x
xi
'
to
xi
with
from
xi
ponds, as
we need only
shown by the
17 SIDES.
39
This corres-
the greater, y v
upper
Upon
we obtain the
we
project from
intercept
+x
2.
t
,
Fig.
we connect
If
this point
7.
we
the inter-
above and
cut off
+ 1 below,
as desired.
x2
If
we
circle
which
axis of X,
lies in
quadratic equation.
We
have
finally to
ratic equation
below.
We
and
z x of the
fourth quad-
above and
A with
-f-
y x below, from
obtaining.
+ 4 above with y
FAMOUS PROBLEMS.
40
upon the
off
desired.
+ 2x = 2y
(y4-4)y
For
yi
Fig.
It cuts off
upon the
The equation
of the line b
2yix
and
its
8.
is
Y4
segment
y4
then
+ (y 4) y = 2y
4
-4
4,
yi
If
we
circle
If
we
X,
we
obtain z 1
= 2 cos
we have only
to
draw a
A perpendicular
first
erected at this
in the
17 SIDES.
41
CHAPTER
V.
We
shall
straight edge
2.
how
is
how
the ancients
We
and the
modern mathematics
Let
it
cubic equation
x
+ a)<2+ bx + c = 0>
or the biquadratic,
x
+ax
bx 2
+ cx + d=0.
ALGEBRAIC CONSTRUCTIONS.
Put
=y
43
+ ay + bx + c =
+ axy + by + ex d = 0.
xy
and
The
-\-
The equation
x
two
conies.
=y
xy
-f-
ay
+ bx + c =
0,
FlG.
section
One
10.
is at infinity
finite distance,
is
the same.
The hyper-
bola (Fig. 11) has again one asymptote parallel to the axis of
is
no longer perpendicular to
this axis.
FAMOUS PROBLEMS.
44
The methods
3.
We
pose.
shall
b.c.)
(Fig.
(in the
on the right)
and the diameter perpendicular to it.
Draw lines from 0, the vertex of the
circle thus determined, to points upon
the tangent, and lay off from
upon
each the segment lying between its
intersection with the circle and the
figure the vertical tangent
tangent.
Fig.
12.
the axis of X.
If
we produce
of*
off
cos 6
circle is cos
hence
0.
The
difference of the
two segments
cos
cos 6
= sin
cos
6'
to the
the circle
is
1,
by the
r,
and
ALGEBRAIC CONSTRUCTIONS.
By
transformation of coordinates
equation,
(x
The curve
we
45
+ y )x-y = 0.
2
is
and
is*
tote.
cir-
cular points.
this curve,
We now
This cuts
off
This
is
of
=1
the segment
X,
and
which
point y
0, x
1, and hence of the line joining this point
to the point of the cissoid.
We
now
see
1,
draw a
Y the
the axis of
x
off
intercept 2,
= 0,
line
and through
from the origin
its
Let
off
upon
to the tangent x
/
Lay
1.
The
inter-
2.
as follows
constructed.
3
A.
(c.
be a fixed point, a
150
its
b.c.) is
constructed
FAMOUS PROBLEMS.
46
If
line.
we
directions a segment
is
b,
off
on
According as b
the conchoid.
and lay
its
is
the origin
jugate point
a,
a node or a con-
is
for b
a it is
through
lei
line,
the fixed
to
we have
r_
'
whence
(x
-fy
)(x-a) 2 -b
The conchoid
fourth
is
= 0..
then of the
has a double
order,
=
+y
the line x
a.
factor (x 2
Further, the
immediate importance.
any angle by means of this curve in the
following manner
Let <
MOY (Fig. 13) be the angle to
ter of
We may
trisect
OM
b,
an arbitrary length.
b describe a circle,
of
X with origin
On
With M
the side
OM
as a center
lay off
and radius
Construct the
ALGEBRAIC CONSTRUCTIONS.
47
degree.
It
r ~3~~
<
<
+4
7r
~~
infinity.
5.
In
all
braic curves,
we must
We
need an instrument which shall trace the curve by a continuous movement, for a construction by points is simply a
method
of approximation.
PART
IL
CHAPTER
I.
Transcendental Numbers.
1.
axis of abscissas.
we
throughout (berall
small there
less, as
is
an
dicht),
infinite
i.e.,
numbers
fill
in
number
of such points.
Neverthe-
of points
upon the
axis
is
way
between
among
made
<o
+ a^-
-f
algebraic numbers.
+ a n_ + a =
l0 )
an algebraic number. Of
Rational numbers
occur as a special case in equations of the form
with integral
course
we
coefficients is called
-|- a.i
0.
FAMOUS PROBLEMS.
50
We
now ask
the question
Does the totality of real
numbers form a continuum, or a discrete series
that other numbers may be inserted in the intervals ?
:
algebraic
such,
X These new
rendus, 1844,
is
is
rather complicated.
The
given by Georg Cantor in a memoir of fundamental importance, lieber eine Eigenschaft des Inbegriffes reeller algebraischer
LXXVII,
making use
of a
1873).
We
more simple
it is
true, sug-
numbered, so to speak.
them occupies a
definite
This proposition
may
be
stated as follows
The manifoldness of real algebraic numbers and the manifoldness of positive integers can be brought into a one-to-one
correspondence.
We
seem here
to
meet a contradiction.
first
can be associated
TRANSCENDENTAL NUMBERS.
infinite masses.
51
example, that we
It is evident, for
may
thus
12
2n
In dealing with infinite masses, the words great and small are
inappropriate.
As a
positive integers.
We
all algebraic
> n + a^"- +
1
+a
_ lW
+a =
n
all
we
defined by
= n-l+|a + |a +
l
ai|
'
{,
+ ,a
|,
as usual.
given number
To a
obviously
finite.
FAMOUS PROBLEMS.
52
|a
|ax
|a 2
|a 3
|a4
Equation.
<KN)
2xl=0
x2=0
-1
+2
12
3x
x2
1
1
x-1=
x - 2 =
-3
-1.61803
z3 =
2x - =0
1=
- 1.41421
3
2
Among
-1
x =0
2
1
Roots.
1
1
-0.70711
- 0.61803
- 0.33333
+ 0.33333
+
+
+
+
+
0.61803
0.70711
1.41421
1.61803
3
reducible,
TRANSCENDENTAL NUMBERS.
N
is
finite
We
53
N only a
is
We now
= 0.999
'
FAMOUS PROBLEMS.
54
in
Take
Cantor's process
is
first five
is
number
as follows.
different
the 6th decimal of the first algebraic number, for 7th decimal
number
different
The proposition
is
then
demonstrated.
We
see
are far
CHAPTER
IL
ir.
ir
shown
knowledge
goes, has
if
ancients
the
is
then impossible.
It
is
extremely interesting to
to find a solution
worked
for
how
advancement
these neces-
i the
mani-
excellent
translation
of the
FAMOUS PROBLEMS.
56
Among
1.
may
measurement or by
distinguish the
first
was
to be attained
by
direct estimation.
oldest
Papyrus
The writer
area.
following rule
The value
circle.
of
2.
thus obtained
Much
very inaccurate.
it
(1
The Greeks
is (-y-)
less accurate
Kings,
7.
23,
is
= 3.16
the value
2 Chronicles,
it
circle
work kvkXov
3,
4. 2).
and
not
fierpyja-Ls,
com-
cir-
it
by Huygens
De
practical
magnitudine
circuit
inventa.
As
effect the
_1
Consider for example the curve y
sin x, which repreGeometrically, ir
sents the sinusoid with axis vertical.
from the
we
Any
ir.
= sin-1
is
called an
THE CONSTRUCTION OF
integral curve because
57
it.
it
/dx
The ancients called such a curve a quadrairix or TerpaywviThe best known is the quadratrix of Dinostratus
which, however, had alb.c.)
350
(p.
ready been constructed by Hippias of
t,ovaa.
Elis
(p.
420
an angle.
b.c.) for
the trisectin of
Geometrically
it
may
be
Having given a
defined as follows.
and
B,
the quadratrix.
From
this definition
and from
follows that y
it
proportional to
is
7T
= we have
1,
It
= lim
for y
tan-y
is
0.
FAMOUS PROBLEMS.
58
_2_
hence
7T
axis of X.
rectifica-
This
we have no apparatus
by continuous
movement.
Fig. 15 gives
ir
or
is
Fig.
less
than
it.
15.
-1
not so convenient as the curve y
sin
x, but
appear that the latter was used by the ancients.
3.
names
analysis.
it
does not
to 1770, characterized
rise of
by the
modern
THE COMPUTATION OF
59
n.
Numerous
3^-5
This same period brings the discovery of the mutual dependence of e and
The number
tt.
e,
first
found in principle in
first to
= cos x +
which, for
x =
e^
becomes
tt,
e in
This formula
mathematics.
is
transcendence of
sin
it,
x,
l.
4.
most remarkable
in all
of the transcendence of
is
e.
rightful place.
Among
its
it
dis-
tt is
transcendence of e
is
here proved.
FAMOUS PROBLEMS.
60
An
analogous proof
Hermite, was
for
{Mathematische Annalen,
ir9
closely
Lindemann
by
given
XX,
1882.
related
to
Ueber die
that
of
Zahl
ir
still
very
complicated.
The
first
simplification
'
But Hurwitz and Gordan soon showed that this transcendental formula could be done away with {Gttinger Nach; all three papers are reproduced
with some extensions in Mathematische Annalen, Vol. XLIII).
The demonstration has now taken a form so elementary
that it seems generally available.
In substance we shall
follow Gordan' s
mode
of treatment.
CHAPTER
The Transcendence
We
1.
III.
of the
Number
e.
well-known series
e
= 1+ i+ 2 + -"^+
i
for
we
easily
that
after
1000th
the
has
x
for
n!
any
finite
value of
when
becomes
infinite
We
now propose
The number e
an algebraic number,
with integral coefficients of the form
is
not
F(e)= Co + C 1 e + C2 e2 +is
impossible.
The
coefficients
i.e.,
+C
Q may be
n e*
an equation
supposed prime to
one another.
We
method
of demonstration, show-
in the
following
!;
FAMOUS PROBLEMS.
62
We
way.
a certain integer
by
so that
MF(e)=MC + MC ie + MC
We
shall
(1)
Each
Me 2
M K and
+MCn e n = 0,
+'
2e
-Me may be
11
a fractional part
eK ,
sepa-
and our
= MCo + MA + M
+ C +C
M F(e)
The
n en
integral part
MC+MA +
is
2 e2
iei
(2)
+ "+ M Cn
+
+ C = 0;
C2
not zero.
-+M
Cn
by a prime number
it
The expression
(3)
Ciej
+C +
2 e2
'
+C
n en
fulfilled,
from
zero,
The
and a proper
sum
is
of an integer different
may
error
we may assume
e,
'
'
en
assumed equation.
2.
We
make use
shall
in our proof of a
symbol
and a
The symbol
Thus,
we
is
= 1 4^ h 4^ h 4^
2
4^
^
h"
4-
form
63
e.
Suppose,
we have given
e.g.,
a developed polynomial
= tc
f(x)
rx
r
.
We
represent by
%c
f(h),
r
>
the
sum
c 1 -l
But
if
f (x)
+c
is
-2!
+c
-3!+-
'
by
+ h) = 2 c
+ h) =
r
r (k
c' r
-n!
and
f (h)
is
to
finally replace
r !.
f(k
the
+c
depending on
The polynomial
c'r
c' r
r!,
k.
<(x)
is
the
* (x)
where p
is
x)(2
A x)-
^
= xp-,[(1
1
a prime number,
and
|C
|,
-Cn
x)p
(p-l)\
and
later
e.
we
We
shall
shall suppose p
make
it
increase
without limit.
To
we
con-
= *.
At the points
1, 2,
in an odd number of
X is tangent without
and n the curve remains
meets
it
For values of
between
recedes indefinitely.
FAMOUS PROBLEMS.
64
Of the function
properties
x being
1.
<(x) tends
values of
Put
<
we
(x)
will
now
toward
its
sum of
the absolute
terms.
= x (1 x) (2 x)
x)
(n
"^
YW
jut
we may then
write
^p
1)!
<f>
r=np+p
^ c,x-=
)=2
7
+(X
r=p-l
w + w
C'XP-1
C''*X
np+P-1
+.
X
(
Hence
r=np+p
*(h)=2
cjtf.
i^p-1
all
1)
!,
which
by
r!,
This
since
p_1
is h
1) !. All the terms
(p
of the development after the first will contain the factor p.
As
to the
first, it
may
be written
(1-2-3-
-n)P-(p-l)!
1)'
Cp
and
is
Therefore
and hence
<(h)
<j>
by p since p
(n!) p
(h) ^z 0.
'
>
(mod.
n.
p),
<
3.
(h
We
<j>
(h) is a
very large,
is
to
T=f^
an integer
have
(h
<
and
1,
n,
divisible by p.
+ k) = %c
+ k) = %
r
(h
a formula in which
term
its last
viz.
being an integer,
k) is
even
65
e.
we
c' r h
are to replace h r
by
h.
r(
_(
One
l-k-h)(2-k-h)---(-h)---(n-k-h)] P
(p-1)!
h in
the development
is
hence the
of degree
p.
r=np+p-l
4>(h
The
2c' h
+ k) = r=P
r
r
.
still
denominator (p
1)!.
may be
( l) k P
1)
written
kP- 1 [(k
(n
- k) !]p
= (<
(h
3,
+ k)
is
We
then divisible by
kP
kP" 1 [(k
- 1)
(n
p.
F<V)=C,-f- C ie
is
impossible.
+Ce
2
+-
+C
n e"
- k)
!]p
p.
FAMOUS PROBLEMS.
66
(h)
=C
F (e)
we
<f>
select
(h) -f
<
Crf
(h), so
that
+C
(h) e
2 </>
(h) e
+C
n <
(h)e*.
(h) e k, into
an
ek
<f>
(h)
=e
%c
r
.
*.h*=h-+
Replacing
by
or
!,
-2T +
Vr + 7+i)T +
'
'
what amounts
to the
of the quantities
rh'"
r(r
1) h'"
r(r
1)
= +j
h*
h'-'k
The
+ k)
fii'st.
r
;
2
,
(r
1)
-2-h,
fractions,
+ '~ ^ h"V +
'
+jhk'"1 +k r
k
k'
.r
(h
+1
+
'
(r.+ l)(r
+ 2)
same form
as the
development of
we have the
series
+ H
k
OH-rr
l-t-(r
less
* = i+ k +
l)(r
+ 2)-rh-
+_+...
3!
,
J<^
line in the
qr,k' e
expansion of e k h r
represented by an expression of the form
The second
k\
may
therefore be
67
e.
sum
%c
it
%c
=%c
^q
%q
this
sum
when
r
,
simply <(h-f
is
Further
p (2, 3).
r)k c r k
r)k c r k
*(k)
a fortiori of
+ ky + e
(h
= 2jc
k);
this is a
(2, 1),
k'j
and
of e
Jq
r)k
c,k
r
,
ek .
consideration,
F(e)<A(h)-C ^(h)
all
the terms,
now easy
terms of the
the
first,
|
<(h) is
get finally
+C
+C
2 e2
It is
we
n en
the
first
(h
+ n)
All the
by p for
by p; hence
are divisible
is less
the
suitable choice of p
is
true of their
sum.
Since an integer not zero and a fraction cannot have zero
for a sum, the
assumed equation
is
e,
impossible.
or Hermite's Theorem,
is
CHAPTER
The Transcendence
1.
7T
The demonstration
given by Lindemann
the case of
IV.
of the
Number
it.
number
e.
is
C + C ie + C 2 e 2 ++ Cn e Q =
cannot exist, Lindemann generalizes this by introducing
-
e2
e,
e
e
kl
11
sums
of the
k2
+ e ++e +
l2
in
form
+ e k*
+ ev
(1)
+C (e + e
kl
k2
+-
+ e k*)
+ e v)+-
=0
+ C^e + e +
where the coefficients Q are integers and the exponents k
11
12
t,
Ij,
of higher order.
* Leibnitz
calls a function
interscendental function.
x\ where X
is
an algebraic
irrational,
an
+ a^' +
ax N
bx
'
+ b^'- +
1
kt
+ bN =
i5
These equations
etc.
terms equal to
69
+ aN =
it.
first
symmetric functions of
our later developments need
1.
not be integers.
will be sufficient to
it
bli,
ak N
*
,
bl ff ., etc.
+ axy*" + a ay- +
+
-f btf*'- + b by
1
y
'
N '-2
+ a^a*- = 0,
1
+ b^b*'" =
1
0, etc.
We
assume equation
members by an integer
such as
M(e kl
M(e
ll
we multiply both
k2
M(ekl +
into
to be true
(1)
+-
+ e kN
),
fraction, thus
+ e +- + eks)=M +
+ e + -+e v)=M +
k2
c1 ,
2 ,
l2
C M
+C
+C
ll
+C M +
+ cl +
8
0.
FAMOUS PROBLEMS.
70
We
shall
suitable choice of
the
sum
of
an integer not
divisible by a certain prime number p, and consequently
that the fractional part can be made as
different from zero
small as we please, and thus we come upon the same contra-
represents
line
first
diction as before.
2.
We
symbol
M =^(h),
=r! and
where
follows
select as
if/(x) is
a gene-
formed as
A(x)^
l 1)!
(p
-x)-
[Oi x)
x)
(l 8
(l w
x)]
b* p
b N p b*" p
'
a prime
is
(k-x)]-a*-a*'P-a*"P-
'
where p
[(k 1 -x)(k 2
of
and
later will be
assumed
the factors a Hp b N p
'
have
in the
aN b N
,,
As
to
development of
i/r(x)
quantities
that
is,
ak 1;
ak,
bli,
bl 2 ,
ak N
bl,,,
Later on
we
shall
have
The presence
we wish
same
each divided by (p
1.
of these
\f/(h)
is
1)
!.
divisible by p
and
con-
xj/
powers of
(h) in increasing
r=Np+N'p+
iKh)
+p
=2
h, it
71
it.
cr h r
r=p-l
In this development
ators
!.
ak x ) p a N p a N
'
_ lv
(ak x
(blx
( l) irp+N
'p
(p-1)!
ak2 bl 2
bl N .) p b
x -1 p N p N "p
a a
Np
bN
"
"
'
'--(a N a
(b H .b
N '-1 p sp
b b*" p
)
limit,
\\i
(x)
=Jc
sum
We may
<Kx)
write
= 2c
= X_ lv [a
x
r
ir
ay
b*b w '(k 1
x)(k x)-
0-x)(l-x)Since for
stant,
x of
we may
-(k x
(U- x)
it
by
K.
We
-x)
is
]'
a con-
then have
(xK)^ 1
* = =)! K
'
1.
FAMOUS PROBLEMS.
72
J ^(k +
The expression
We
h) is
an
of
(x)
is
integer divisible by p.
have
-k -h)-k,-h)(l 2 -k ,-h)-
a-'Pb-P[(l 1
The vth
-(-h)-
t/
(l w
-(k N
-k,-h)]P
-k ,-h)p
J
line is
if/
h,
r=xp+x'p+
Kk
+ h)=
h is h p
+p-l
c'M
r=p
whence
v=y
r=xp+N'p+
2^(k +
h)
+p
C' r h
r
.
F=p
l'=l
The numerators
and integral,
symmetric functions of the quantities
ak L
ak 2
bli,
bl 2 ,
ak x
-,
bl N ,
Similarly for
v=\
We
if/
We
3.
now
n.
equation
(1)
+C
(e
kl
+e k2 +-
/f(h),
+ e k -)+C
(e
ll
l2
+e +-
-e
v )+- =0
cannot be true.
by
thus obtaining
k
+ e k^(h)] +
=0,
= +
=+
ek
The product
ek
tf,
(h)
= ek % c h = % c
may
r
.
tk + + l)(r + 2) +
(r
of every
term of the
series
+ + l + + l)(r + 2) +
(r
is less
the series
k
Hence
r
q rk being a
than
<e' k
+ 1 + + l)(r + 2)
+1
'
(r
'
or
(r
'
+ l)(r + 2)
=-n
.Pl k
qr, k e
'
1.
We may
e
'
then write
^(h)= 2 c
k r
e h
=Jc
(h
+ k) + % c
r
q r>k k
e'
kl
is
less
FAMOUS PROBLEMS.
74
By
giving
snm
ing the
and form-
n,
V(h) +
V(h) +
="SV (k,
+el ^(h)
+ h) +T je W 2 cJVfcJ-
all
Co A (h) 4-
dig* (k, +
h)
v=l
+ Ci$te\K\ Clk
p=l
By
2,
qT K
,
+ C^V
+ h) +
v=l
(| F
+ c/s'S
v=l
>=0.
as small as
we
please
by taking
p sufficiently great.
Since q rk
|
<
1,
a fortiori
of
r
and hence
also of
2 2
?=1
c r k ,q r)k e \K\
r
of (2) can,
after the
by p
ible
is
but the
first
(1).
We now
The proof for the more general case where C = may be reduced
by multiplication by a suitable factor, or may be obtained directly
to this
by a proper modification
of
\p
(h).
75
it.
quence of the latter. For this reason we shall call it Lindemann's corollary.
The number e cannot satisfy an equation of the form
kl
ll
C'o+C' 1 e +C' 2 e +-
(3)
lf
'
'
this, let k 2 , k 3
when
the exponents
To demonstrate
Form
<>,
kx
kK
similarly for
may
2,
3,
etc.
>,
be deduced from
(3)
JC'o+C' 1 e*+C' 2 e
/3
= Co + Ci (e + e +
+C
(e
k2
kl
kl+ll
+e
he
kl+l2
a
1
k
*)
+C
1, 2,
=l,2,-
-,*
(e
kl+k2
+ ek2+ka +
+' )+
i?
Our product
demonstrated.
We may now
deduce a
still
satisfy
where the
=0
algebraic numbers.
(2)
C ?,
(
C\
FAMOUS PROBLEMS.
76
If
we multiply
we
get
the product
= l,2,-
+ C ?ek + C ?V+
0=1,2,
a, jB.-y,
= C + C k e +C e ++ C e k+ k +C e +
+
+
k
1, 2,
-,!
n2
k>k
where the
coefficients
+-
>
the quantities
c<>>,
qj>,
cv,
c<;>,
c<?>,
c<?>,
By
rational.
the previous
we have
proof such an
.
accordingly Linde-
+C e +C e
k
k,
I,
+-
Ci,
This
may
C
the exponents
5.
and
+C e +C e
k
coefficients
From Lindemann's
of interesting results.
+-
corollary
we may deduce
a number
an
For consider the remarkable equa-
immediate consequence.
tion
1
-f-
irr
0.
not.
\ir is
Therefore,
hence the
transcendental.
it is
6.
1 =
coefficients
exponent
77
it.
=e
-We know
x
.
that
e.
transcendence of
This
e.
is
We
= =
=
=
i.e.,
all
points
exponential curve
is
A further
= sin-1 x
and
The function
sense, of
the
is
the
function
similar functions.
y
= sin -1 x
2
ix
is
= e e~
iy
iy
.
We
We may
The curve y
= sin -1
x,
=e
0, y
x
,
passes through
0.
CHAPTER
V.
of
it.
of
it,
is
tt is
the
An
actual construction of
ir
we must use
If such a
construction
is
continuous motion.
2.
Such an apparatus is the integraph, recently invented
and described by a Russian engineer, Abdank-Abakanowicz,
and constructed by Coradi of Zurich.
Y=F(x)=/f(x)dx
when we have given the
differential curve
= f(x).
For a
description of this
original
mempir
ingenious instrument
(in
we
refer
;
the
fuller
to
the
in French,
GEOMETRIC CONSTRUCTION OF
79
it.
fig.
to
the point
(x, y).
ig.
The apparatus should be so constructed then that the tracmove parallel to the variable direction of this
This
is
effected
The
construction of
Take
ir is
is
of especial interest.
the approximate
presses this
roller.
3.
A weight
+y2==r2
FAMOUS PROBLEMS.
80
the integral curve
is
then
=/ Vr 2
dx
= - sin- - + Vr - x
2
2
.
it
r7r
o,
Upon
the lines X
7T
2
r
If
we make
The
1,
->
4
3 7T
-t-j
2
r
7r
or its multiples.
instead of a pencil.
circle
curve
is
by them.
e\e*^
Bo may
r Srni
TITLE
4^s?>
VK,
1]
'
Mm
IM
iHHI
PMBili
1118111111]
'
':'
"
:
:^i!!iiiji';i;i:! :liiVi;;