FIBONACCI SPACES

E.D. Cashwell and C.J. Everett

University of C a l i f o r n i a , L o s Alamos Scientific Laboratory
L o s Alamos, New Mexico

..

INTRODUCTION
The Fibonacci sequence -j F . j = 0, 1, 1, 2, . . .
, with F . +
w
J
J
F. . = F - + 2 J J - > m a y be r e g a r d e d as one e l e m e n t of a c e r t a i n
space of s e q u e n c e s a s s o c i a t e d with the q u a d r a t i c polynomial f(x) =

2
- 1 - x + x , from which its r e m a r k a b l e p r o p e r t i e s d e r i v e . In the
following p a g e s , we p r e s e n t first, i n . m o d e r n a l g e b r a i c terminology,
an exposition of those p a r t s of the g e n e r a l t h e o r y of such s p a c e s of
linear r e c u r r i n g s e q u e n c e s which form a background for this point of
view.

The s p a c e s a r i s i n g from q u a d r a t i c polynomials a r e then con-

s i d e r e d in this setting, with some applications to n u m b e r theory, in

p a r t i c u l a r to v a r i o u s t e s t s for p r i m a l i t y of the M e r s e n n e and F e r m a t
numbers.
It is hoped that the p a p e r m a y thus s e r v e a s an introduction and
s o u r c e of r e f e r e n c e for t h e s e a s p e c t s of the subject, [ l ]
1.
THE SPACE OF A POLYNOMIAL
Let

f(x) = - a 0 - a , x - . . . - a

,x

+x = (x - r , ) . . . ( x - r )

an a r b i t r a r y monic polynomial of d e g r e e n in Z [ x | ,
coefficients

a. in the domain

with

Z of r a t i o n a l i n t e g e r s , its (complex)

z e r o s being t h e r e f o r e a l g e b r a i c i n t e g e r s .
set C(f) of all s e q u e n c e s

i.e.,

be

With f(x) we a s s o c i a t e the

S = i s , s , . . . ^ with components

the complex field [ 2 ] C, in which

s , ...,s

s. in

a r e a r b i t r a r y but

having
(R)
x
'
for all j > 0.

a_s. + a 1 s . , 1 + . . . + a
, s.
. = s.
0 j
1 j+1
n-1 j+n-1
j+n
C l e a r l y , C(f) is a v e c t o r space of o r d e r n over

G.

An obvious b a s i s c o n s i s t s of the i n t e g r a l s e q u e n c e s [ 3 ] ( i . e . , with

components in Z):
U =
= 1

0 Kji f
0,a0,...j .....Un.1={un_lF.J={0....,l,aii_1,
''Worked p e r f o r m e d under the a u s p i c e s of the United States Atomic
Energy Commission.
**Refer to footnotes at end of a r t i c l e .
97

98
of

FIBONACCI SPACES
C(f), i n t e r m s of w h i c h e v e r y s e q u e n c e

April

S of t h e s p a c e m a y be e x -

p r e s s e d u n i q u e l y , in t h e f o r m
S = s n U n +0 . . + s
,U
,, s.C
0 0
n-1 n-1
j
(
2
)
A " G e o m e t r i c " sequence /jl,r, r , . . . / with
a n d o n l y if f(r) = 0 T h u s
R

I1' ri' ri ' ' * ) '

c o n s t i t u t e t h e g e o m e t r i c s e q u e n c e s of
c,R,
1 1

r ,C i s i n

C(f).
,

C(f), i n p a r t i c u l a r t h e i n t e g r a l

(R); t h e s e , t o g e t h e r w i t h t h e l e s s o b v i o u s
(a

)m + a
n-m

are "Newton's

v. = r . J +
j
1
satisfy the r e -

recursions

' v , +. . . + a
,v
, = v , 1 < m < n
n-m+1 1
n-1 m - I
m
~"

sequences

if a n d o n l y if t h e z e r o s of

R=f"r.],

with

formulas".

The g e o m e t r i c
C

if

Every element

s e q u e n c e V = R . +. . . + R = 1 v . I - \ n, a
, , . . . ; ,
.
1
n
I jj
I,
n-1
)
. . . + r , j > 0.
B e i n g i n C(f), i t s c o m p o n e n t s v
lations

C(f)

' i = 1, . . , n

+. . . + c R , c . C
n n I

of t h e s p a c e t h e y s p a n i s t h e r e f o r e in

R.

f(x)

a l s o f o r m a b a s i s for

are distinct.

i = l , . . . , n ; j = 0,...,

n - 1 ,

For,

C(f)

over

the m a t r i x

h a s the V a n d e r m o n d e

deter-

minant

TT,
.(r, - r . ) . When the l a t t e r is not z e r o , the i n v e r s e m a -1 k r " ^
" r ^A"
-1
trix R
= j r . . J e x i s t s [_4J ( o v e r C) w i t h I = R R, f r o m w h i c h i t
follows that
(B)

U. = r . - R . +. . . + r . R ,
i
il 1
m n

By m e a n s

of t h e s e e q u a t i o n s ,

i = 0, . . . ,

n - 1

w h i c h m a y be r e g a r d e d a s t h e " B i n e t

f o r m u l a s " for the g e n e r a l c a s e , e v e r y s e q u e n c e

S

= (SjHS0U0+Sn-lUn-lj

is e x p r e s s i b l e uniquely in the f o r m
S=c1R,+..,+ c R ,
1 1
n n
with

s . = c , r . J +. . . + c r J , Jj > 0,
j
11
n n
'

c. C
I

C(f)

1966

FIBONACCI SPACES

when the r.

a r e distinct-

99

This underlying s t r u c t u r e of the s e q u e n c e s

of C(f), p a r t i c u l a r l y of the i n t e g r a l s e q u e n c e s , is one of the m o s t

c u r i o u s f e a t u r e s of the subject.
F o r e x a m p l e , if all z e r o s of f(x) a r e h - t h r o o t s of unity, it
is c l e a r that e v e r y S = I c . R .

is pure p e r i o d i c , with period | 5 J di-

viding h. When a / 0 and the r. a r e distinct, the e x i s t e n c e of a

sequence S, with all the c. d 0, of period k, i m p l i e s that all r.
are
l

'

k - t h r o o t s of unity. F o r , a / 0 i n s u r e s , via (R), that S is p u r e

i k
p e r i o d i c , and we should then have a linear s y3 s t e m I r . f r . - l ) c . = 0.
l

j = 0, . . , n - 1, with d e t e r m i n a n t det R d 0*

Thus

(r.

'

for each i, and e i t h e r

c. = 0 or
l

r.

- ljc. = 0

is a k - t h root of unity,,

Con-

sequently, if f(x) has distinct z e r o s which a r e r o o t s of unity, and h

i s the l e a s t positive i n t e g e r for which all r.
S=

Ic.R.

to h.

For,

of C(f) w i t h a l l

= 1, then e v e r y

c. / 0 is p u r e p e r i o d i c of period k equal

by the f i r s t r e m a r k ,

k|h,

and by the second, h k,

h e n c e , e v e r y period k = h9
Note:

The following g e n e r a l i z a t i o n of the " g e o m e t r i c b a s i s "

t h e o r e m was suggested by the r e f e r e e :

m for

f(x), then r is a z e r o of the f i r s t

F r o m this one m a y show that the m

s h) = (

a r e in C(f), w h e r e
for

If r is a z e r o of multiplicity

h. > 1.

} rJ;

m - 1 d e r i v a t i v e s of f(x)

sequences

5 h=

O'1-*"-1 -

(^) = 1 for h = 0, and 0

Moreover,

if

1 i k 1 n, of m u l t i p l i c i t i e s

f(x)

= (j)(j-l). . . ( j - h + l ) / h !

has the distinct z e r o s

m ,.-.., m

r , , . . . , r, ,

r e s p e c t i v e l y , then the set of

n s e q u e n c e s c o n s i s t i n g of k s u b s e t s such a s that above, one for each

zero

r., a r e l i n e a r l y independent,

and so form a b a s i s for

C(f)

in

the g e n e r a l c a s e , provided only that z e r o , if it o c c u r s among the

has m u l t i p l i c i t y 1.

r.,

The linear independence follows from the non-

s i n g u l a r i t y of the "confluent Vandermonde matrix,, " Cf refe [ l " | .

2B
If lim s
S of C(f),

then

LIMIT THEOREMS

/ s . = r e x i s t s for a t e r m i n a l l y n o n - z e r o sequence
r

m u s t be one of the z e r o s of

f(x)*

For,

the

100

FIBONACCI SPACES

April

recursion (R)
imp Lies a A + a 1 ( s . 1 1 / s . ) +. . .+ a . (s . . , / s .) = s . , / s . .
N
'
^
0
1 J+l J
n-1 j+n-1' y
j+n' j
Since

s . . 0 / s . = (s . , _/s . . , )(s . , . / s .), and so on, we have

f(r) = 0 in

the limit. As a partial answer to the questions arising here we include

Theorem l . ( A ) .

If one zero

r,

of f(x) exceeds all others in

then S= i s . > = I c . R . has lim s . / r . J = cn

I ])
i i
J 1
1
c, 4 0, S is terminally non-zero, and lim s. , . / s . = r . .
1
j+1' j
1

absolute value,
if

(B).

Hence

| r . | < 1 for all i 2! 2, the sequence S has lim (s . - c , r, k ')= 0 o

If

i
Hence, if S is an integral sequence, s. is the closest integer to
for large

j , and, if a, ^ 0

same c, .
Proof.
(A).
If c, / 0, then

n o

other integral sequence

c,r-

I c . R . hats the

s . / r . J = c, + c JVr , / r , )J +. . . + c (r / r , ) J ~* c, .
j' 1
1
2 2' 1'
n rr 1
1
s. / 0, j 2 J, and so s. , / s . =
r

( s

j +

i/ri

J + 1

)/(s/riJ^rici/ci

= r

(B).
s. - c , r / = c^r J +. . . + c r J^ 0o Twote integral sequences with
N
'
j
1 1
2 2
nn
^
the same c, a r e therefore terminally identical, and hence identical,
if a n ^ 0. Indeed, it is clear from (R) that two sequences a r e equal if
they agree on any n consecutive corresponding indices.
3.

INTEGRAL SEQUENCES

The integral sequences

with integral basis

F = -If. I of C(f) form a module

U
' .., U

Z(f),

- , every such sequence being uniquely

expressible, with integral coefficients, in the form

F = f n U n +. . . + f _U , ,
0 0
n-1 n-1
The sequences of

Z(f) with f

f.6Z .
j

= 0 form a sub-module of Z(f), and

have remarkable divisibility p r o p e r t i e s , as we shall see.

The sequences of Z(f), considered modulo m, form a finite module
ring

Z
Z

(f) of exactly m

sequences with components in the modular

= Z/Zm, the first

n a r b i t r a r y , the r e s t governed by the r e -

cursion (R) mod m.

Suppose

now that m ^ 2 is an integer prime to

F = jf. ( be a sequence of

a ,

(f). It is clear from (R) that

and let

F is pure

1966

FIBONACCI SPACES

101

p e r i o d i c if it is p e r i o d i c at all, hence that F is p e r i o d i c if and only

if i t s n - t u p l e s
(f0.

.... f ^ )

and (fk,

...,fk+n.1)

a r e identical for some positive k, the l e a s t such k being its p e r i o d .

Since

F = 0, of p e r i o d k = 1, is the only sequence of

taining the z e r o n - t u p l e , a non- z e r o sequence

m
different
11
k 4 m - 1.

n-tuples,

Moreover,
m

(f)

con-

F can contain at m o s t

and so m u s t be p u r e p e r i o d i c of period

F has period k = m

- 1 if and only if its f i r s t

- 1 n-tuples

(f0> . . . . t^.ify
a r e all distinct.
quences of
this p e r i o d .

. . , . f n ), . - . . (f m n_ 2 , . . . . f m n + n _ 3 )

In such a c a s e ,

each of the m

- 1 non-zero se-

(f) is a t e r m i n a l s u b - s e q u e n c e of F, and so a l s o h a s
The situation cannot a r i s e for a c o m p o s i t e modulus

m = ab, a .> 2, b > 2.

For,

such an F contains the n - t u p l e s ( 1 , 0 ,

. . . , 0) and (b, 0, . . . , 0), hence

a F is in the space and contains the

n - t u p l e s (a, 0, . . . , 0) ^ (0, 0, . . , 0), which is i m p o s s i b l e .

The m a x i m u m period m

- 1 i s attainable in c a s e of a p r i m e

m o d u l u s , which m a y be s e e n from the t h e o r y of Galois fields [?]

Let p be a p r i m e in Z, and suppose f(x) = -3L -SL, X - . . . - a

-x

+x

is an i r r e d u c i b l e

polynomial in Z [ x ] .

Such an f(x) e x i s t s for

e v e r y p and n.

The s e q u e n c e s of Z (f) m a y then be r e g a r d e d not

only as the i n t e g r a l s e q u e n c e s , mod p, of the space

a quite different light.

F o r t h e r e e x i s t s a field

field" of f(x) ( a b s t r a c t l y , the Galois field

e l e m e n t s , uniquely e x p r e s s i b l e in the form
n

C(f), but a l s o in

C* O Z , the " r o o t n
n
GFp )) of exactly p

s = j 0L + f ,1r + . . . + fn-1, r " ', fj. Z p

in t e r m s of which we m a y w r i t e
f(x) = (x - r)(x - r P ) . . . (x - r P
with n d i s t i n c t z e r o s in

C*.

n-1
)

102

FIBONACCI SPACES

April

Following the plan of g 1, we find that the set C*(f)

of all s e -

quences

S = ] s . j , with s . C * governed by the r e c u r s i o n (R) in C*,

^
n n
is a vector space of o r d e r n over C#, consisting of exactly (p )
s e q u e n c e s , and containing the o r i g i n a l module Z (f) of p
sequences
p
)
.( , f . Z .
|J
J
P
The z e r o s of f(x) being distinct in C*, the g e o m e t r i c s e q u e n c e s
R. which they g e n e r a t e form a b a s i s for C*(f), in t e r m s of which e v e r y
sequence of the space m a y be uniquely e x p r e s s e d :

S=c,R, +..,+ c R ,
11
n n

c. C*
j

with components s . = c, r . J +. . . + c r J , iJ > 0. w h e r e the r. a r e now

^
j
11
n n ' - '
i
the z e r o s of f(x) displayed above.
This s u b - s t r u c t u r e of the s e q u e n c e s
character.

F o r , the m u l t i p l i c a t i v e group G of C* is cyclic, with a

generator
h|p

S reveals their periodic

s of p e r i o d p

- 1, and the e l e m e n t

- 1.
rP

If h is the period of r in G, then

h a s p e r i o d h / ( h , p ) = h, which is t h e r e -

fore the p e r i o d of e v e r y z e r o of f(x)0

An obvious modification of the

a r g u m e n t at the end of i 1 shows that e v e r y jion- z e r o sequence S of

C#(f) h a s p e r i o d

h.

The e l e m e n t r itself need not have period p - 1; however, some

e l e m e n t s ji 0 of C* does g e n e r a t e the full group G, and its m i n i m a l
polynomial mod p is i r r e d u c i b l e of d e g r e e n in Z [ x ] . Hence t h e r e
e x i s t s , for e v e r y p and n, a n i r r e d u c i b l e f(x) of d e g r e e n in Z [ x j
for which the z e r o r in C# g e n e r a t e s G, and e v e r y sequence S / 0
of the c o r r e s p o n d i n g space C*(f) has p e r i o d p - 1. We s u m m a r i z e
t h e s e r e s u l t s in m o r e conventional t e r m s in
T h e o r e m 2.

If m ^ 2 is an i n t e g e r p r i m e to a n , then, modulo

m, e v e r y i n t e g r a l sequence

F = if. {. of C(f) is p u r e p e r i o d i c of p e r -

iod k < m - 1. Hence, if m |f , then a l s o m | f , f , f , . . . . If

m is c o m p o s i t e , e v e r y p e r i o d is l e s s than m - 1. If p is a p r i m e ,
and

f(x)

-^C ft*
r
i s i r r e d u c i b l e |_6J m o d p ,

all integral sequences

a r e p u r e p e r i o d i c w i t h t h e s a m e p e r i o d h, w h e r e
p

and

n, t h e r e e x i s t s a n
E x a m p l e 1.

For

f(x)

such that h = p
3
f(x) = - 2 - x + x m o d 3,

h|p

- 1.

of

C(f)

For every

- 1.

V = | 0 0 2 0 2 1 2 2 1 0 2 2 2 0 0 1 0 1 2 1 1 2 0 1 1 1 0 0 , . . .] ,

1966

FIBONACCI SPACES

*~#C

103

the p e r i o d [ ? ] being n = 26 0
Ifaninteger
quence

m .> 1 divides some

F , the l e a s t positive such j is called the r a n k of m in F 9

C o r o l l a r y 1.
f

If

is- an i n t e g r a l sequence of

= 0, e v e r y positive i n t e g e r m

-1 in

p r i m e to a

C(f),

with

has a r a n k r <^ k < m

F.
4.

THE SPACE OF A. QUADRATIC

2

F o r f(x) = - a n - a..x + x
C(f)

f., j _> 1, of an i n t e g r a l s e -

of all s e q u e n c e s

(R)

= ( x - r ] )(x-r ), the a s s o c i a t e d space

S = -j s. [ , s . C ,

Vj + a l S j+l

= S

satisfying

j+2'

J=

has the b a s i s U n = j 1, 0, a , > , U, = J09 1, a , . . . j ., and contains

the s u b - s p a c e of v e c t o r s

c.R
1.

i = 1,2

+ C-R., w h e r e R. = $ l , r . , r . , e o e ,
L

LL

L*

a r e its g e o m e t r i c s e q u e n c e s .
V =R

I + R 2 = h j = | 2 ' a i> 2a o

consisting of the i n t e g e r s

v. = r

The p a r t i c u l a r sequence
+a 2

i ' ---j

+ r ? , j > 0, is of s p e c i a l i m p o r t -

ance.
The g e o m e t r i c s e q u e n c e s

R. form a b a s i s for

if r ] ^ r , in which c a s e the m a t r i x R
1
r

C(f) if and only

of d 1 i s

r2

rr2

1
-1
The c o r r e s p o n d i n g "Binet f o r m u l a s " a r e a c c o r d i n g l y
(B)

UQ = ( - r 2 R x + r 1 R 2 ) / ( r 1 - r^
Ul = (R x . R 2 ) / ( r 1 - r 2 )

or explicitly,
u

0j

= (

"r2rlJ

*\vzV^\-TZ>>

lj

= (r

lJ "

2J)/<rrr2)

The r e l a t i o n u~. = a^u^ . . is h e r e m a n i f e s t .

0j
0 l,j-l
We know from g2 that, if | r | > | r | , l i m u . . / r
lim v . / r ,l
3

= 1, and lim u, . , - / u , . = lim v. ,. / v . = r_ .

l,j+l
lj
J+l J
1

=l/(r,-r

)}

104

FIBONACCI SPACES
Again,

if

| r ? | < 1,

then,

April

for e x a m p l e ,

sequences

) = 0

and

v . i s t h e n e a r e s t i n t e g e r t o r , , j 2: J; m o r e o v e r - i f
only i n t e g r a l s e q u e n c e R, + c ? R .
The i n t e g r a l

l i m (v. - r

F = If. j- of

C(f)

a n f 0, V i s t h e

form

a module,

with

integral basis

U n , U , , a n d , m o d u l o a n i n t e g e r m p r i m e to a , a r e
u
i
-p

u
all p u r e periodic with p e r i o d s l e s s than m . F o r a p r i m e modulus p,

if f(x) i s i r r e d u c i b l e m o d p , t h e s e p e r i o d s a r e e q u a l , a n d d i v i d e p - 1,
2
f(x) s u c h t h a t e v e r y p e r i o d i s e x a c t l y p - 1.

and t h e r e exists an

The s u b - m o d u l e
the single b a s i c

of s e q u e n c e s

sequence

F = f, U,

(i. e . , w i t h

f = 0) h a s

U1 , w h i c h i s h e r e a f t e r d e n o t e d s i m p l y b y

U = j u . ( = < 0, 1, a, , . . { .

For

f(x) = - 1 - x + x , i t i s of c o u r s e

the

Fibonacci sequence.
2
Every integer
where

prime

k i s t h e p e r i o d of

to

U mod m; indeed

s t a t e m e n t s m a y be m a d e f o r e v e r y
U alone.

v. = a u.
E x a m p l e 2.

+u

in

U,

F = fnUn.

C(f)

is e x p r e s s i b l e in

V = 2LL + a U

follows

j > 1

For

f(x) = T - 2x + x 2 = (x - 1 ) 2 , U =
E x a m p l e 3.

s e q u e n c e of

For example, from

r < k < rn

m |u, , u ? , , . . . ; s i m i l a r

F ~ f, U, a n d

It i s i n t e r e s t i n g t h a t e v e r y
t e r m s of

has a rank

0,1,2,3,4,...

. V. ^

2,2,2,2,2,..

For

f(x) = 2 - 3x + xZ= ( x - 2 ) ( x - 1), u . = 2 j - 1, v . = 2 j + 1

s a t i s f y t h e s i m p l e r e c u r s i o n s u . ,, = 2 u . + 1 , v . . . = 2 v . - 1. N o t e t h a t
J
J+1
J
p
9k
J+1
F
u = 2 - 1, v T = 2^ + 1 .
The s e q u e n c e U m o d u. h a s p e r i o d j .
P
2k
J
E x a m p l e 4.
For
f(x) = - 2 - x + x 2 = (x - 2)(x + 1), u
f o r odd

p, and

v ^ =2^
5.

= (2P + l ) / 3 , v

= 2P - 1 ,

+ 1, k 2: 1.
THE SEQUENCE

E v e n f o r t h e g e n e r a l q u a d r a t i c [8]',* 'the s e q u e n c e

= JO, 1, a

U = -]0, 1, a 1l 3, a Q + a x

2
,

1966

FIBONACCI SPACES

105

has some remarkable properties, which flow from the

Lemma 1.

For all

j i o, t > o, a 0 u.u t + u j + 1 u t + 1 = u. +t+1 . .

The statement is easily proved,
from

j , j+1

to

j+2,

being obvious for

for fixed

t 2 0, by induction

j = 0, 1.

The induction step

reads
a

0 U j+2 U t +U j+3 U t+l

a

= a

0 ( a 0 U j + a l U jHhl ) u t

0 U j+t+l

+a U

l j+t+2

+ (a

= U

j+t+3

0 U j+l

+a U

l j+2)ut+l

'

From this follows

Theorem 3.
I
implies
Jj k
F
'

i.e..

Since

The correspondence ju. preserves divisibility

I
I
^
u. u. , or u. u., u n ., u 0 ., . . .
J' k
j
j
2j
3j

u n = 0, u

= 1, the final statement is trivial for

j = 0, 1.

Fixing j 2. 2, we prove u. |u.

by induction on q > 1. This is trivial
for q = 1. Fix q 2. 1, and assume u. |u. . Setting t = jq - 1 ( > 1)
j JH.
u., . , v .
j j(q+i)

Lemma 2.
p divides
If

If a prime

p divides any two consecutive

u , , then

an

p |u. , 1 , u. j

but not

a n , thenfrom

(R)

follows

p [u., u. ,

and ultimately P|u n = 0, u.. = 1, which is false e

Theorem 4.
Let m be a positive integer prime to a n . Then,
2
I
modulo m, U is pure periodic of period k < m , and m \M , u^,
uon , . . .
e Thus m has a rank r <_ k in U.
Moreover, m |u ,
2k
.
U
u , u^ , . . . and no other u., i . e . , m u if and only if r \JL . In
r
2r
j
<
particular, r |k.
We have only to prove
for

.- 0.

Write

m |u

For JL >. 1, we have

Jl - rq + j , 0 < j

< r, q 2 1.

implies
r <. ,
For

r \ , which is obvious

by the minimality of

r.

t = rq - 1 (2 0), Lemma 1

reads
a^u.u
. +u.,,u
= u^
0 j rq-1
J+1 rq
*
Since
is rprime to

m \u
u , u
(Th. 3), we have m a^u.u
, . Now m
1 a>
X.
r
rq
' 0 j rq-1
a^, and hence also to u
, , since a prime
common to
r
0
rq-1

106

FIBONACCI SPACES

this and m is also in u

m |u., where

0 < j < r.

April

, contradicting Lemma 20

Hence

j = 0 and = rq.

It follows that

We turn to the spe-

cial case of
Lemma 3.

If

(a n , a ) = 1, the sequence

(a)

(a Q ,u.) = 1, for all j > 1,

(b)

(u.,u

(c)

(uo u k ) = ( u k ' u i + k )

Proof.

U has the properties:

) = 1, for all j > 0,

f o r a I 1

J k * -

(a) Induction on j 2. 1. For

(a n , u. , ) = 1 for fixed

j = 1, trivial.

j > 0, we see from (R) that (a , u.

Assuming

?)

For, a prime common to these divides

a^u.,., and hence

1 j+1
(a , a,) = 1, violating the induction hypothesis.

- 1 also.

u.,,,
j+1

since

(b) is clear from Lemma 2 and (a).

(c) is trivial when j

or

k is zero.

For j , k ^ 1, we have from Lemma 1, a^u.u. : + u. ,, u, = u. .. . CI e a r I

P J k~1
J+1 k
J+k
ly (u. u k ) |u k , u
and hence (u , u k ) | ( u k , u . + k ) . Conversely,
(u, , u. ,, ) a n u.u.
The former is prime to a
vides

by (a) and to u,

by (b).

u., u,k , and so it divides (u., u1k ).

J
J
Note:
It is clear from (a) that the only integers

components

u. of

Thus it d i m

dividing

U are prime to a .

Theorem 5(A).
with g n

If Jg-f is an. a r b i t r a r y sequence of intege r s

= 0, then the correspondence j--> g. p r e s e r v e s g. c.d. l s , that

is- (gj?'S k ) = g(^,k)

(B)

ifandonlyif

(g , gfc) = (gfc, g.+k)

In particular, the sequence

for all j , k > 0.

U has this property whenever

(aQ, a x ) = 1.
Proof:
we may suppose

(A)

The necessity is obvious. Inprovingthe

| k.

sufficiency

The conclusionis clear for k = 0 , since g n = 0.

If 2 k> 0 and k \jg = qk, we have

1966

FIBONACCI SPACES
y

*u,

k) = %

<%' gk>

= (g

107

k ' g2k> = = <8 k ' g]

In the r e m a i n i n g c a s e , . > k > 0, k j. , we w r i t e

< k <-^, and obtain (g , gfc) = ( g ^ g
c l e a r how the E u c l i d e a n a l g o r i t h m ,

- kq + j , 0 < j

) = . . . = (gfc, g ).

It is then

p r o c e e d i n g from this r e l a t i o n

through a sequence of s i m i l a r s t e p s and t e r m i n a t i n g in an equation

such a s

L = KQ + J, 0 < J < K < L, with {, k) = J | K , leads to the

conclusion

(g^ , g k ) = . . . = ( g ^ , g R ) - (g K ,

(B)

gj)

gj

The application of (A) to U is now c l e a r from L e m m a

3(c).
Note:

The n o n - t r i v i a l p a r t of Th. 4 follows elegantly from Th,

5(B) when (a n , a ) = 1. F o r , if r is the rank of m , and m \Uf , then

m

l ( u r > uxi;) = u \/ r,
xny)

C o r o l l a r y 2.
correspondence

l(r,.i?)|r,

If r , / r

j - u. = (r

and r = (r,)\X

a r e r e l a t i v e l y p r i m e i n t e g e r s the

- r ? ) / ( r , - r ) is g. c, d. - p r e s e r v i n g .

F o r , f(x) = (x - r , ) ( x - r ? ) = - a n ~ a x+x
an = - r r

has a, = r , + r?

and

relatively prime.

Note:

It i s . . c l e a r that the set of all g. c. d. - p r e s e r v i n g func-

tions g(j) on the i n t e g e r s is a closed a s s o c i a t i v e s y s t e m ( s e m i - g r o u p )

with identity under the composition f(g(j)).

T h e o r e m 5(A), suggested

by the r e f e r e e , c h a r a c t e r i z e s t h e s e functions.

The s e q u e n c e s

U re-

sulting from q u a d r a t i c s with (a , a ) = 1 a r e n o n - t r i v i a l functions of

the kind.

As a " t r i v i a l " e x a m p l e c o n s i d e r

g(2j - 1) = l , g ( 2 j ) = 2j

Although "well-known" we include the seldom fully stated

C o r o l l a r y 3.
ference

d = r

For integers

- r , let u. = (r

(a)

(d,u.) = (d,j)

(b)

Aprime

p |u

ifandonlyif

r, ^ r ?
J

with (r , r ? ) = 1 and dif-

- r ? ) / ( r , - r ), j 1.

p |d.

Such a p r i m e

p has rank p

in U.
(c)

If p j d , then p | u

and (d, u ) = 1.

(d)

If p | d , then p |u , (d, u ) = p, and if p is odd, p ju

108
(e)

Every prime factor

(f)

If r
<1

Proof.
Uj

April

FIBONACCI SPACES

> r

q ^ p of u

is of form. 1 + hp.
P
> 0, then for every odd prime p there exists a prime

l(p) dividing u s> of form 1 + hp, and q(p) is one-one.

P
(a) is trivial for j = 1, and follows for j > 2 from

= (d+r 2 ) J - r 2 j ) / d = (d j " 2 + C J 1 d j - 3 r 2 + . . . + C j . _ 2 r 2 J " 2 ) d + jr 2 J '" 1 (a*)

since

(d,r?) = 1 .
(b) From (a*), u

of p follows from (a).

- dp

mod p.

The statement about the rank

(c) follows from (b) and (a).

(d) follows from (b), (a), and the congruence
2
p

implied by (a*) for a prime

1 or

p; and q f r , , r ? .

= r

mod q, r, ^ r

mod q, since

We present two proofs:

(1)

^ 1 mod q implies

n-1

= r_p~
pi

there exists a prime

+ . . . + r_ P
2

q = q(p) of form

> pr
r

plest being 2, 1 with u

r , , r~

= 2-1.

mod q,
otherwise

> p, it follows that

=

1 + hp 7 dividing u

odd, and by Cor. 2, this function is one-one.

tion is valid for every pair

p - per-

n-1

iod (r 1 rJL mod q) |<I>(q) = q - 1 = hp. (2) r^

= 1 = r^
q
, , u ) = u( ,
), by
Hence rp (q - 1) = hp,
n i (u
J Cor. 2.
q-1 p
'q-l,p
'
(q - l,p) = 1, q | u 1 = 1.

Letting

Lt

r ? r ' = 1 mod q, we have ( r , r ' ) P = 1, r . r '

(f) Since u

mod

p ^ 3.

(e) If q |u , q / p , then r
(d, u ) is

u =. pr
P

for

Of course the construc-

covered by the corollary, the simIt is not known whether an infinity of

primes 1 + 2p exist.
Corollary 4
If
(a 0 , a,) = 1 and a , a

> 1, then u

=0 <u

=1 < u =a

< u^ < u . . . . ,

and

(a) j composite implies

except in the single case

a'

u. composite, (b) u. |u, implies j | k ,

J
J K
= 1. If a, = 1, (a) is false only when

1966

FIBONACCI SPACES

109

j = 4 and u^ is p r i m e , while (b) is false only when j = 2 and k is

oddo

Proof.

It is c l e a r from (R) that u. is i n c r e a s i n g as stated.

(a)
Let j = hi, h 2 2, i > 2.
By Th. 3, u h , Ujju., w h e r e
u, , u. < u., since 2 S h, i < j . If u, or u. exceeds 1, u. is c o m n

p o s i t e . Suppose both a r e 1. Then c l e a r l y h = i = 2, and

1 = u, = u.

= u ? = a:, , j = hi'= 4, u. = u ,, and (a) follows with its p r o v i s o .

(b)

If u. |u , then u. = (u., u ) = u , . ,* (Th. 5), w h e r e

(h k) ^ j . If equality holds, j jk; if inequality, we m u s t have (j, k) =

1, j = 2, 1 = u, = u "= a . , k odd, and (b) follows.
60

THE LUCASIAN SEQUENCES U,. V

The i n t e g r a l s e q u e n c e s
U = {u.J"= { 0 , 1 , a , . . . ]

and V = {v J = [2, a 1 ? 2 a ( ) + a 1

,...];

with
u . = ( r 2 J - r 2 J ) / ( r r r 2 ) , v. = r ^ + ^
of the space

C(f) a s s o c i a t e d with the q u a d r a t i c

f(x) = -a Q - a x x + x

= (x - r ^ x - r 2 )

where
r

(a

i i

+ Q2)J r

(a

2 i i"

I
Q2)

'

=a

i 2 + 4 a o'

1_
r j + r^ = a 1 ? r ^

= -a Q , i^ - r^ = Q ,

aQQ / 0

have c u r i o u s i n t e r r e l a t i o n s , which have been exploited by Lucas [ 4 ;

p . 2 2 3 ] , and L e h m e r | " 5 j , (in even m o r e g e n e r a l form) in the design
of v a r i o u s " t e s t s for p r i m a l i t y " .

We p r e s e n t h e r e some old and new

a s p e c t s of t h i s .
The following r e l a t i o n s a r e e a s i l y verified using the above fornix
u l a s for

u., v.:
3 J

110

FIBONACCI SPACES
+

(la)

(1)

Yk

<2>

u.vk - ukv. = -2(-a0)juk_.

( k > j)

(3)

v.vk + Qu.uk = 2v.+k

(3a) v . 2 + Qu. 2 =

(4)

v.vk - Qu.uk = 2(-a0)jvk_.

(k > j)

\VV

April

JVU2J

Z^.

(4a) v . 2 - Qu. 2 = 4 ( - a Q ) j
(5)

v 2 . = v . 2 - 2(-a 0 )J

(6)

u x = 1, v 1 = a]L

(7)

=Q

'

mod p, for e v e r y odd p r i m e

p.

F o r e x a m p l e , we compute
1_
P

1_

= ( a 1 + Q ) P - (a x - Q Z ) P = X P ( l - ( - l ) V ^ ^ O

Z Q u

ZP~lu p = i P0 c P I - 1a P ^ "
i odd
u

1_

'/

from which

, Q ( p - 1 ) / 2 m o d pF .

A p r i m e is said to be r e g u l a r ( r e l a t i v e to
We know from Th. 4 such a p r i m e has r a n k r
U, w h e r e k(p) i s the p e r i o d of U mod p, and
no other u.. More r e m a r k a b l y , we see now,
(7), that

f(x)) in c a s e p | 2 a n Q .
= r(p) |k(p) <. p - 1 in
p | u n , u , u 7 , . . . and
from (2), (1), (6), and

(p Tl 2
^ mod p^ .
1 - Q ^ ' ^ ^ a , 1 = 02 a n u p-1. and a,1 + Q " ^ a 1 1 = 2u p+1

L e m m a 4.
w h e r e p |u

If p is a r e g u l a r p r i m e , then p |u

if (Q/p) = 1,

if (Q/p) = -1 (Legendre symboll ), so that always

r(p) < p + 1 in U.

1966

FIBONACCI SPACES

111

L e m m a 5.
If p is a r e g u l a r p r i m e , then p | v
k+1
if r(p) = 2
in U.

if and only

Proof.
If p l v - k then p |u 2 k+l b y ( l a ) > b u t p T u ? k b Y ( 4a )
hence r(p) = 2
. C o n v e r s e l y , if p ju-^+i but not u ^, then p |v k
by ( l a ) .
These a r e the b a s i c l e m m a s ,

F o r computational r e a s o n s , we

note that the sequence ^?v k i , with v ? = 2a n + a,

, v?]z+i - ( v ?k)

- 2 a n ^ , k 1 (cf. [ 5 ] ) is r e l a t e d to the a u x i l i a r y sequence | S-, |

defined by
S x =. 2 + ( a i 2 / a 0 ) , S k + 1 = S k 2 - 2, k > 1

^k- 1
via the s i m p l e equation v -^ = a n
S- , k > 1.
Thus, w h e n e v e r
a n ja, , 7S j is an i n t e g r a l sequence, and a r e g u l a r p r i m e p divides v?k- if and only if p |S, .
We m a y state one of L e h m e r ' s r e s u l t s as
T h e o r e m 6.
suppose a n , a,
(a)

Let M = 2 - 1 , w h e r e q is an odd p r i m e ,

and

a r e i n t e g e r s with the p r o p e r t i e s

If p is a p r i m e d i v i s o r of M, then p | 2 a 0 Q w h e r e

Q = a,

4a Q .
(b)

M prime implies
Then

(Q/M) = -1 and (a /M) = 1.

is p r i m e if and only if

I v /iyr+i \/2

(equivalently

S q _ 1 , if a j a ^ ) .
Proof.
If p r i m e P I M I v / M + i \ / 2 ' P i s r e g u l a r b Y ( a ) ' hence,
in U, r(p) = M+l <, p+1 by L e m m a s 5, 4. Thus p | M < p, and M = p,
prime.

Conversely,

if M is p r i m e , it is r e g u l a r by (a), and from

(b), (3), (6), and (7) follows

a]_

- Q = 2 v M + 1 , or v M + 1 = -2a Q mod

M. Then from (5) and (b), -2a Q = ( v / M + 1 ) / 2 ) 2 - 2 a Q ( M + i ) / 2 , and M

divides the stated v..
3
Example 5.
For,

For

a n = a, -2,

Q = 12, (a) and (b) a r e satisfied.

only the p r i m e s 2 and 3 divide

2a n Q, and M = 1 mod 2 and

M =. 1 mod 3. M o r e o v e r , if M is a p r i m e (Q/M) = (3/M) = - 1 , and

112

FIBONACCI SPACES

April

(a /M) = (2/M) = 1, since M =. -1 mod 8. Since a n | a 1

, the sequence

\ S, J is i n t e g r a l with S, = 4, and M is p r i m e if and only if M |S

,.

In the s a m e fashion one m a y p r o v e

T h e o r e m 7

Let F = 2^ + 1 , t 2 1, and suppose a , a,

are

i n t e g e r s with the p r o p e r t i e s
(a)

If p is a p r i m e d i v i s o r of F, then

(b)

F prime implies

p]2anQ0

(Q/F) = 1, ( a Q / F ) = - 1 .

Then F is p r i m e if and only if F v,

if a

z^v

,.

/oi

(equivalently

oK }-

Proof.

Suppose p r i m e p | F | v / T r i\/7

Then p is r e g u l a r and

in U has r(p) = F - 1 < p + 1, so p | F S p + 2, C l e a r l y F = p, o t h e r wise

2p <. F < p + 2, p < 2. If F is p r i m e , it is r e g u l a r by (a), and

from (b), (6), (7), (4),

F,

a,

- Q = -2aQv

F r o m (5) and (b), 2 = (v / T r i \ / o )

mod F or v

, = 2 mod

+ 2 mod F and the t h e o r e m fol-

lows.
Example 6.

For

an = a

= 3, Q = 2 1 , only the p r i m e s 2, 3, a n d

7 divide 2 a n Q , w h e r e a s F = 1 mod 2, F = 2 mod 3; as for 7, note

t
t
that e i t h e r 2 = 1 + 3h, and then F = 3 mod 7, or 2 = 2 + 3h, and
then F = 5 mod 7. Hence (a) holds Q If F is p r i m e , we have ( 3 / F )
= (F/3) = (2/3) = - 1 , and a l s o ( 7 / F ) = ( F / 7 ) = - 1 , since
(5/7). Hence
quired.

The a u x i l i a r y S,

Note:

(3/7) = -1 =

(Q/F) = ( 3 / F ) ( 7 / F ) = 1, and ( a Q / F ) = ( 3 / F ) = - 1 , as r e a r e i n t e g e r s , with S, = 5.

The t e s t s indicated in Th. 7 have no computational ad-

vantage over the orthodox N and S condition 3

''

= -1 mod F

Indeed, the l a t t e r is a s p e c i a l c a s e of Th. 7, with a n = 3, a, = 2.

the l a t e s t computational

Comp Q 18 (Jan. 1964) and Scientific A m e r i c a n (Nov. 1964, p. 12).

l e a s t undecided F e r m a t n u m b e r is
7.
Let

For

r e s u l t s see the r e l e v a n t a r t i c l e s in Math.

F

THE SPACE

The

.
Z (f)
P

2
be a p r i m e of Z, and f(x) = - a n - a , x + x ^ Z [ X U

The

1966

FIBONACCI SPACES

113

r e g u l a r i t y condition p J 2 a Q, which we h e r e a s s u m e , i n s u r e s that p

is odd, aand the z e r o s
distinct.

r.

of f(x) in its root field a r e n o n - z e r o and

2
2
Since f(x) = ( x - 2 ' a , ) - (2 ! ) Q, w h e r e 21 is the i n v e r s e of 2
mod p, we see that this root field is Z if (Q/p) = 1, or the Galois
field C* = GF(p 2 ) of
4 3 if (Q/p) = - 1 .
o

(I.)
If (Q/p) = 1, t h e r e is a b 6 Z such that b = Q mod p,
and f(x) is reducible in Z [ x |. . Indeed, in Z , f(x) = (x - 2 , (a 1 +b))
p

(x - 2'(a, - b)) h a s the d i s t i n c t z e r o s indicated.

itself has b a s i s

R , R

over

The space

Z (f)

Z , each of its s e q u e n c e s being exJ

p r e s s i b l e in the form

s. = c , r , + c ~ r ~ , c , r. Z . Since r.
=1
j
11
22
i i p
i
mod p, e v e r y S is p u r e p e r i o d i c of p e r i o d dividing p - 1 . By an a r gument now f a m i l i a r , if h is the l e a s t exponent for which both r.

= 1,

e v e r y sequence with both c. p 0 (e. g. , U n , U, and V) h a s period

h.

?t

This c a s e is of s p e c i a l i n t e r e s t when p = 2^ + 1 is a p r i m e .
The conditions (Q/p) = 1, (a /p) = -1 then hold if and only if f(x)
has z e r o s in Z with (say) r
a q u a d r a t i c r e s i d u e , and r
a nonp

r e s i d u e . In such a c a s e , r
h a s period p - 1 , and r ? m a y have any
J
z
ll
v p - l ) j in
period
m
U
(
Z
.
F
r
o
m
the
above
e
x
p
r
e
s
s
i
o
n
of. s.J in t e r m s
ii2 ^
P
of the r. it is c l e a r that one sequence (S = 0) has p e r i o d 1, exactly
p-1

(those with c, = 0 , c

/ 0) have period m, and the

maining s e q u e n c e s (with c, / 0 , c
(II.)
r, r
r

If (Q/p) = - 1 , f(x) is i r r e d u c i b l e in Z [ x ] , with z e r o s

in C*, w h e r e , in the cyclic group G of o r d e r

have a p e r i o d h | p

- 1.

quences

S/0
(r)P

of C*(f),
+1

- (^f

+l

- 1, r

The g e o m e t r i c s e q u e n c e s

and form a b a s i s for the l a t t e r space over

Since

re-

a r b i t r a r y ) have period p - 1 .

in C*(f)
h.

(p-l)p

in p a r t i c u l a r those of
= rP

+1

and

R. a r e now
C*.

All s e -

Z (f), have period

( l - r P " l ) = 0, p f u p + r

F o r e v e r y p > 3, t h e r e e x i s t s an i r r e d u c i b l e q u a d r a t i c ( r e l a 2
tive to which p is n e c e s s a r i l y r e g u l a r ) for which r h a s p e r i o d p - 1 .
2
E v e r y sequence of the c o r r e s p o n d i n g space has p e r i o d p - 1 . Since
p |u , , and e v e r y pair (0,1), . . , (0, p - 1 ) , indeed e v e r y pair / (0, 0),

114

FIBONACCI SPACES

April

m u s t a p p e a r exactly once as an adjacent p a i r in the sequence fu n , . . . ,

u 2 i|

mod p, it is c l e a r that p+1 is itself the r a n k of p in U, and

the above sequence c o n s i s t s of the t e r m i n a l e l e m e n t and p-1 blocks of

p+1 e l e m e n t s each.

Moreover,

each block a r i s e s from the f i r s t by

r e n a m i n g its e l e m e n t s , since each is the beginning of a sequence of the

space which is a multiple of U itself.
Such a sequence thus p r o v i d e s a solution of v e r y s p e c i a l type for
the m = p, n = 2 p r o b l e m (Cf. footnote [ 7 ] ) . L e h m e r ' s q u a d r a t i c (Ex.
2
3
""
2
5) f(x) = -2-2x+x mod p = 2 - 1 = 7 has a root r of p e r i o d 7 - 1 = 48,
and mod 7 we find
U = {01262216 05323352 04131143 06515561 02454425 03646634 0..J
FOOTNOTES
1.

Most of the ideas p r e s e n t e d m a y be found e l s e w h e r e ,

in l e s s g e n e r a l f o r m .

2.

r o o t f i e l d of
30

f(x)

f(x)

o v e r the r a t i o n a l field,

mod p.

o r by the a b s t r a c t

S e e [7~|.
the

u . = a u
, . , , j 2l 1
oj
o n-1,j-1

Explicitly, for

i = 0, 1, . . . , n - 1 ; j = 1, . . . , n,
r

w h e r e the
gree n - l - i

= ( - U M n - l - i i j ) / TV (r - r )
k^j K J
a denotes the e l e m e n t a r y s y m m e t r i c function of d e 3

in the n-1

when n - l - i = 0).
5.

C i s e v e r y w h e r e r e p l a c e d by t h e

A l t h o u g h l i n e a r l y i n d e p e n d e n t , one m a y n o t e , a m o n g o t h e r s ,
relation

4.

See for example r e f e r e n c e s [ 3 , 6, 101.

A p a r a l l e l v e r s i o n i s o b t a i n e d if
" r o o t f i e l d " of

sometimes

roots

Cf. ref.

other than r . .

(Here <r = 1

[8].

P e r i o d always m e a n s m i n i m a l period, while pure periodic m e a n s

t h a t p e r i o d i c i t y o b t a i n s f r o m t h e b e g i n n i n g of t h e s e q u e n c e .

The r o o t field for a r e d u c i b l e

properties are more

7o

f m o d p e x i s t s , but the p e r i o d i c i t y

complicated.

T h e m e t h o d i n d i c a t e d ( w i t h s u i t a b l e i n s e r t i o n of a z e r o ) p r o v i d e s

FIBONACCI SPACES

115

an a l g e b r a i c c o n s t r u c t i o n of a sequence of i n t e g e r s mod m, of
length m + n - l ,
prime m.

containing no r e p e a t e d n - t u p l e , in the c a s e of

The e x i s t e n c e of such s e q u e n c e s for a r b i t r a r y m , n

is a well-known c o r o l l a r y of a t h e o r e m on g r a p h s [ 2 , 9~L

(Re-

m a r k of r e f e r e e . )
F o r the Fibonacci c a s e , see f 1 0 ] , on which the p r e s e n t section
is modelledo
REFERENCES
J. W. Archbold, A l g e b r a , 1958, P i t m a n , London, p 0 419.
C. B e r g e , Theory of G r a p h s , 1962, Wiley, N. Y. , p. 167, Th.
2.
E. B. Dynkin,

P r o b l e m s in the t h e o r y of n u m b e r s , Survey of

r e c e n t E a s t E u r o p e a n m a t h , l i t e r a t u r e , 1963.
G. H. Hardy, E, M. Wright, Introduction to the t h e o r y of n u m b e r s , 3rd e d ' n . , 1954, Oxford, p . 223.
Do H. L e h m e r , An extended t h e o r y of Lucas functions, Annals
of Math. (2) 31, 1930, 4 1 9 - 4 4 8 .

On L u c a s ' s t e s t for the p r i -

m a l i t y of M e r s e n n e ' s n u m b e r s ,

J o u r . London Math. Soc. , X,

1935, 162-165.
E. L u c a s , Theorie des n o m b r e s , 1961, B l a n c h a r d , P a r i s , Ch.
XVII, XVIII.
C. C0 MacDuffee, Introduction to a b s t r a c t a l g e b r a , 1940, Wiley,
No Y 0 , p . 174 ff.
L. M i r s k y , An introduction to l i n e a r a l g e b r a , 1955, Oxford, p.
36.
O. O r e , T h e o r y of g r a p h s , Am. Math. Soc. Colloq. P u b .
XXXVIII, 1962, p . 40, Th. 3 . 1 . 3 .
No N. Vorob'ev, Fibonacci n u m b e r s , P o p u l a r l e c t u r e s in m a t h e m a t i c s , v. 2. , Blaisdell, N. Y.
xxxxxxxxxxxxxxx

RELATIVELY PRIME SEQUENCE SOLUTIONS OF NON-LINEAR DIFFERENCE EQUATIONS

J.W. Layman
V i r g i n i a Polytechnic Institute

In the following we p r e s e n t a o n e - p a r a m e t e r family of f i r s t

o r d e r n o n - l i n e a r difference equations which a r e shown to p o s s e s s
solutions which a r e s e q u e n c e s of ( p a i r - w i s e ) r e l a t i v e l y p r i m e n u m bers.

In p a r t i c u l a r , we show that the s e q u e n c e s

= 22

(A)

(B)

G, = 2 cosh
k

where n

+1
k
,-1
29 cosh

0
-j-

1 3

is an odd or even i n t e g e r a c c o r d i n g as the sign in (B) is

plus or m i n u s , r e s p e c t i v e l y , c o n s i s t of r e l a t i v e l y p r i m e n u m b e r s by
virtue of being solutions of such difference equations.

Sequence (A)

is the famous sequence of s o - c a l l e d F e r m a t n u m b e r s originally conj e c t u r e d by F e r m a t to c o n s i s t only of p r i m e n u m b e r s (later d i s p r o v e d

byEuler).

The p r e s e n t method p r o v i d e s a n e w proof of the r e l a t i v e l y

p r i m a c y of t h e s e n u m b e r s (see [ l ] > p. 14).

Sequence (B) is a p p a r -

ently new and is the only other i m p o r t a n t solution which has been obtained in closed f o r m .
These specific r e s u l t s a r e based on the following s i m p l e t h e o r e m :
Theorem:

Ifthe sequence of i n t e g e r s i u, I s a t i s f i e s the difference

equation
(i)

u k + 1 = uj; - bu k + b

where
u
b0

, u
2

Proof.

k = 0, 1, 2, . . .

and

b is integral,

then the distinct e l e m e n t s

h a v e n o c o m m o n d i v i s o r e x c e p t p o s s i b l y f o r t h e d i v i s o r s of
L

We m a y w r i t e equation (1) in the form

u

k"

b = u

k-i

(u

k-i - b )

which by i t e r a t i o n m a y be e x p r e s s e d a s
116

(Continued on page 152.)

AN ALMOST LINEAR RECURRENCE

Donald E. Knuth
C a l i f . Institute of Technology, Pasadena, C a l i f .

A general linear recurrence with constant coefficients has the

form
U

= a

l*

= a

"'"

r-l

= a

u = b, u , + b~u 0 + . . . + b u
,
n
1 n-1
2 n-2
r n-r

n > r

The Fibonacci sequence is the simplest non-trivial case.

Consider,

however, the following sequence:

a)

^ n = *n-l + ^ [ n / 2 J '

n >0

In this case, successive terms are formed from the previous one byadding the term "halfway back" in the sequence.

This recurrence,

which may be considered as a new kind of generalization of the Fibonacci sequence, has a number of interesting properties which we will
examine here.
The sequence begins 1/2, 4, 6, 10, 14, 20, 26, 36, . . . .

It is easy

to see that all terms except the first are even, and furthermore 0
n
2k-1
2k
is divisible by 4 if and only if n = 2
(mod 2 ) for some k > 1.
We leave it to the reader to discover' further arithmetic properties
of the sequence.
The sequence
tion:

<

<$>

has an interesting combinatorial interpreta-

is precisely the number of partitions of the number

2n into

powers of 2. For example, 6 = 4 + 2 = 4 + 1 + 1 = 2 + 2 + 2 = 2 + 2 +

1 + 1 = 2 + 1 + 1 + 1 + 1 = 1 + 1 + 1 + 1 + 1 + 1 , and
ify this interpretation,

let

<t>

= 6.

To ver-

P(m) be the number of partitions of m

The preparation of this paper was supported, in part, by NSF grant

GP-212. Acknowledgement is also made to the Burroughs Corporation
for the use of a B5000 computer.
117

118

AN ALMOST LINEAR RECURRENCE

into powers of 2. If 2n = a + a + 0 . + a , where

April

a 1 _\ a

and each a. is a power of 2, there a r e two cases:

(i) a = 1; then

a1 +. . . + a
. . + a, /2

> . . A a-i
K.

is a partition of 2n-l; (ii) a > 1; then a, /Z + a ? / 2 +

is a partition of n.

Conversely, all partitions of 2n a r e

obtained from partitions of 2n~l and n in this way, so P(2n) = P(2n-1)

+ P(n).

We also find

P(2n+1) = P(2n) by a similar argument; here

only case (i) can a r i s e since

2n + 1 is an odd number.

These r e c u r -

rence relations for P, together with P(l) = 1 and P(2) = 2, establish

the fact that <t> = P(2n).
n
The same sequence also a r i s e s in other ways; the author first
noticed it in connection with the solution of the recurrence relation
(la)

M(0) = 0
M(n) = n + min

(2M(k)+M(n-l-k))

0 4k< n
for which it can be shown that M(n) - M(n-l) = m if 0
< 2n < <> , ,
N
x
'
'
m^rn+1,
and

^V^^n-L^Zn-ll
Recurrences such as (la) occur in the study of dynamic programming
problems, and they will be the subject of another paper.
Let us begin our analysis of <t>
elementary properties.
(2)

by noticing some of its most

By applying the rule (1) repeatedly, we find

*2n+1 = 2(*0+...

n)

Another immediate consequence of (1) is

(x 3 )'

4>l
M * ,2n-1T = < *n2 .
9
2n ' <*>2n+l
The sequence <t> grows fairly rapidly; for example,
4

= 1981471878

*IOOOO = 2 In fact, we now show that

<t>

1 4 X 1

grows more rapidly than any power of n:

1966

AN A L M O S T LINEAR R E C U R R E N C E

T h e o r e m 1.
F o r any power
k
0 > n
f o r a l l n > N, .
n
k

119

k, t h e r e i s a n i n t e g e r

N, s u c h t h a t
k

k+1
*
1 k+1
N be s u c h t h a t (2
+1) > (2 + ^ )
, and let
a =
mm
(0 / n
) .
n
N . 4 n 4 2N
k+1
T h e n by i n d u c t i o n 0
A an
f o r a l l n A N, s i n c e t h i s i s t r u e f o r

Proof:

Let

N 4 n < 2N, a n d if

n > 2N

* n = *n-l

+<

^[n/2] ^ t * - 1 ) * * 1 +j>/2]k+1)

^ //
i \ k + 1 . ,11-Lk+l
1
nk+l
> a((n-l)
+ (-j-)
) = a(l + - E T r ) ( n - l )
1 . k + 1 . 1 N k+l
k+1
(n-1)
= an
r)
n-1
If w e c h o o s e N, $> 1 / a a n d N , >

>
1 xk+l.
1 N k+l
- a(l + ^ )
(n-1)

.
,_ ^
> a(l +

K.

N, t h e p r o o f i s c o m p l e t e .

K.

We n o w c o n s i d e r t h e g e n e r a t i n g f u n c t i o n f o r 0 .
&

n
(4)

F(x) = 0

+ 0 x + 0 2 x 2 + <*> x

Let

+ ...

Notice that
( l + x ) ( F ( x 2 ) = <

+ 0Qx +0

lX

+ ^ > 1 x 3 + 0 2 x 4 + <*>2x5 + . . .

+ (0r0o)x+(02-01)x2 + (03-02)x3 + (04-03)x4+.

= (l-x)F(x)

thus

F(x) = |E
F(x2) = ill2Hl4- F(x4)
X X
'

(l-x)(l-x*)

We h a v e t h e r e f o r e

l*\
\D)

VM - ( l + x ) ( l + x 2 ) ( l + x 4 ) ( l + x 8 ) - - r
vx/ 7
x

(i-x)(i-x A )(i-x*)(i-x)---

_
~

1 _ _ _ _ _ _
2
g

(i-Kr(i-x)(\-x)(i-x)-

AN ALMOST LINEAR RECURRENCE

120

F r o m this form of the g e n e r a t i n g function,

c o n v e r g e s for

April

we see that

| x | < 1. (As a function of the complex v a r i a b l e

F(x)

z, F(z)

h a s the unit c i r c l e as a n a t u r a l boundary. ) It follows that

lim sup

\/<&

1 >

i. e. the sequence <t> grows m o r e slowly than a

a > 10

for any constant

This is in m a r k e d c o n t r a s t to l i n e a r r e c u r r e n c e s such as the

Fibonacci numbers0
In the r e m a i n d e r of this p a p e r we will d e t e r m i n e the t r u e r a t e of
growth of the sequence

<t> ; it will be proved by e l e m e n t a r y m e t h o d s

that
1
In <t> * -j

(6)

<f>n = e

2
-j- (In n)

In 4

'

4(lnn)2+o((lnn)2)

The techniques a r e s i m i l a r to o t h e r s which have been u s e d for d e t e r mining the o r d e r of magnitude of the p a r t i t i o n function (see F2J ).
We s t a r t by o b s e r v i n g that
^
k
In F(x) = - l n ( l - x ) + V ^ ( - l n ( l - x 2 ))
k=0

EvE E
r=l

k=0

k
2? r

r=l

and hence by differentiation

oo

^-E"'r=l

,+

JS^

2-2-.A 2 '
k=0

r=l
xk~l+.

= 2 + 4x + 2x 2 + 8x 3 + 2x 4 + 4x5+. ..+$,
k

1966

AN ALMOST LINEAR R E C U R R E N C E

where

i s t w i c e t h e h i g h e s t p o w e r of 2 d i v i d i n g

121
k.

Therefore

J_bJ (x)
^ - ( 1 - X ) ( 2 + 6 X + 8 X 2 + 1 6 X 3 + 1 8 X 4 + 2 2 X 5 + . . . + ^^ kx k " X + 8

. )'

w h e r e if

k = 2

a,
1

a
+...+ 2

t h e c o e f f i c i e n t of x

k-1

, a , > a > . . . > a > 0,

1
Z
r -

i n t h e p o w e r s e r i e s on t h e r i g h t h a n d s i d e i s
a,

^
=0 + 0
^k
1 2

+ .. . + e

= a, 2
k
1

+. . . + a 2
r

+ 2k

( T h e r e a d e r w i l l find t h e v e r i f i c a t i o n of t h i s l a t t e r f o r m u l a a n i n t e r e s t i n g e x e r c i s e in t h e u s e of t h e b i n a r y s y s t e m . )
t h e m a g n i t u d e of w,

a -1
0

We c a n e s t i m a t e

as follows:

k + 2k - (2

a -2
+ 2 2

"a '

.
l'

a +1
= (a +2)k - 2

+ a

+ 2 > (1+log k ) k - 2k

hence
(7)

k l o g 2 k - k 4 i | i k 4 k l o g 2 k + 2k

T h i s e s t i m a t e a n d t h e m o n o t o n i c i t y of <t> a r e t h e o n l y f a c t s a b o u t
F(x)

'which a r e u s e d i n t h e d e r i v a t i o n b e l o w .
E-*<ln(l-x))2

Let
G(x) = e
Then
G'(x) _ - l o g ( l - x ) _
G(x) _ In 2 ( 1 - x )

Since the d e r i v a t i v e

X +

5
2 In 2

13
3 In 2

77
1 2 In 2

4
- >

- l o g ( l - x ) / ( l -x) is ( l - l o g ( l - x ) ) / ( l - x ) ,
we
k-1
find t h a t the. c o e f f i c i e n t of x
i n t h e p o w e r s e r i e s on t h e r i g h t i s

where

of

1
'THT

122

AN ALMOST LINEAR RECURRENCE

(9)

hk=l+A-

...

April

Since L = In k + 0(1), we have t h e r e f o r e e s t a b l i s h e d the equations

k=l

k=l

and
(ii)

- x

+ o(k)

This suggests a p o s s i b l e r e l a t i o n between the coefficients of F(x)

those of G(x).

and

Note that if
|^)=(l-x)f(x)

then
x
F(x) = exp

J (l-t)f(t)dt

0
T h e r e f o r e the following l e m m a shows how r e l a t i o n s (10) and (11) might
be applied to our p r o b l e m :
L e m m a 1.

Let
x
A(x) = exp

/
0

(l-t)a(t)dt

(l-t)b(t)dt

x
B(x) = exp

where
A(x) =

Ax

, a(x)

2}k xk_1 '

B(x)

53kxk-

b(x)

^}y*k~l

A s s u m e the coefficients of A(x) and of b(x) a r e non-negative and nondecreasing.

k, A k >

Bk.

Then if a, <_ b

for all k, A, <^ B ; if a, > b

for all

1966

AN ALMOST LINEAR RECURRENCE

Proof:

12:

A Q = B Q = 1. A s s u m e a k < b R for all k, and A < B

0 L k < n,
nA

fo

Then since A'(x) = (1-x)a(x)A(x), we have

n

= a A + a . x(A -A ) + . . . + a. v(A . - A J)7

n o
n-l
1 o'
l n-1
n-2
^< b nA o + b n - l. x( A1. - A o') + . . . + b .l v(A n-1,-A n-20 )7
- A (B -b J + A ^ b
_-b J + . . . + A
.b,
o n n-1
1 n-1 n - 2
n-1 1
< B (b - b .) + B. v(b . -b 0 ) + . . . + B . b, = nB
o n n-1
l n-1 n - 2 '
n-1 1
n

E s s e n t i a l l y the s a m e a r g u m e n t w o r k s if a

^ b

for all ke

The p r o b l e m is now one of e s t i m a t i n g the coefficients of

I^ln^l-x)
G(x) = e
T h e o r e m 20

If
U.

(12)

JL1JL

I J. - J S . J

'V

11

we have
(13)

Proof:

= a In

n + 0((ln n)(ln In n))

Fi-r s t w e show that

00

(14)

In

(1-x) = >
H
x
Zm^ n
m, n
n=m

where
H

=y

^* 1
m-1
a, , . . . , a
. such that 1 = a. < n, and
te
1
m-1
l
This follows inductively, since the d e r i v a t i v e

s u m m e d over all i n t e g e r s
the a. a r e d i s t i n c t .
I

of (14) is

I n ^ V - x ) _ Y"
(x-1)

" ZJ

x11"1

m3nX

April

AN ALMOST LINEAR RECURRENCE

124

and we have
(15)
x

H
= H
, +" ^
m, n
m,n-l
n-1

.
.
m-l,n-l

Turning to equation (12), we have

oo

(16)

oo

2m

V
c x =V
a in (l-x) = ^
X ) ^ ( ) H 7
m
i
/ ^ n
Z^

^"^
^^
ml
n
2m, n
n=0
m=0
n=0 m=0
(We define H
= 0 if m > n, so the p a r e n t h e s i z e d s u m m a t i o n is
m, n
actually a finite ,sum for any fixed value of n. )
Our t h e o r e m r e l i e s on the e s t i m a t e s
. - h
.)m_1 < H
L h1*1".1, if m < n .
n-1
m-1
~
m , n ~ n-1
The righthand inequality is obvious, since this is the sum

(17)

(h

Z^ai'*' m-1
without the r e s t r i c t i o n that the a ' s

a r e distinct.

On the other hand,

given any t e r m of
n-1

xn-1

L.-J a , a
,
,
/
1
m-1
m L a 1 4- n

we form a t e r m
b 1- - - b m -11
belonging to H
, w h e r e b, = a, - r if a, is the r - t h l a r g e s t of
&
6
te
m,n
k
k
k
ja,,,..^
, | o Thus, we d e c r e a s e the l a r g e s t e l e m e n t b y l , the
I 1
m-l.f
second l a r g e s t by 2, and so on; in c a s e of t i e s , an a r b i t r a r y o r d e r is
taken.

No two t e r m s
1
a."* a
1
m-1

m a p into the s a m e

1966

AN ALMOST LINEAR R E C U R R E N C E

125

and
1

m-1

m-1

m-1

s o t h e l e f t h a n d s i d e of (17) i s e s t a b l i s h e d .
P u t t i n g t h e r i g h t h a n d s i d e of (17) i n t o (16), w e o b t a i n
oo
__

(18)

JiiL m - L 2 m - 2
T"^1 a
h . -.

2 ah

c ^ V ^ H ,
n

n Z_^ ( m - l ) I
m =0
On t h e o t h e r h a n d ,

<
2m,n

(19)

> n
n

m
, aM>
( m - 1 )I

2-.
A

*
m =l

I Tv

=e

(m-l).

. 2
ah
n

"

H9
2m, n

f o r a n y p a r t i c u l a r v a l u e of m
We c h o o s e m to be a p p r o x i m a t e l y
2
ah
+ 1, a s s u m i n g n i s l a r g e . T h e n w e e v a l u a t e t h e l o g a r i t h m of
t h e t e r m on t h e r i g h t , u s i n g S t i r l i n g ' s a p p r o x i m a t i o n a n d t h e left h a n d
s i d e of (17),
i
In c

a n d d i s c a r d i n g t e r m s of o r d e r l e s s t h a n

(In n ) ( l n In n):

, / 2a a
,
,2m-11
/n
> In I T
rrr(h
i " ho
i)
/
n
I n
(m-l)I x n-1
2m-l
/
2
2
2
2
= ah
, In a + 2 a h
. ln(h
, -h0
. ) - ah
, (ln(ah
, ) - 1) + O(lnn)
n-1
n-1
n-1
2m-l'
n-lx
n-1'
= a h 2 1 + 2 a h 2 . l n ( l - 2 m " 1 ) + 0 ( l n n)
n-1
n-1
h
, '
n-1
= a h 2 . - 2a h
.h
. + 0 ( l n n)
n-1
n - 1 02 m - l

T h i s t o g e t h e r w i t h (18) e s t a b l i s h e s t h e o r e m 2 0
T h e o r e m 3.

Let

b e a s i n t h e o r e m 2.
,.
hm

Proof:

by (16).

Since

n+1

.
= 1 .

cn
n*-+~ oo
H
., > H
, we have
m , n+1 m , n
n+1
c
n

n
n+1

Then

126

AN ALMOST LINEAR RECURRENCE

April

We also observe that H

< h .H
,
and hence by (15)
m, n "" n-1 m - 1 , n
H
^ <H
+ H l L h .H
;
0
m, n+1 - m, n
n
n-1 m - Z , n
thus

m=l
.

a.

( 2m

3 ah

< --n-c +

~ n+1 n
Corollary 3.

2a

Wr)nZm,n

ml

n-1

2m-l)(2(m-l)
l)!7m-2/v
n
'

(n+1) n-1 m=2

^i I 7

-ps

n+1
If P(x)

V"^ a

n
is any polynomial, and if
y^

n=

ealn

( l - x ) + P(x)

then
In C = In c +0(1)
n
n
P(x)

Theorem 4.
Proof:
n>

.
.
2
- aQ + a x x + a 2 x

In

C
a C
n
0 n + alCn-l
=
c
c
n
n
1
2
0 ^ -p^-(ln n)
n
In 4

Let > 0 be given.

N, (1-)X

< {p

< (I + )

b(x)= ^
We find

<t>
<C
n "~ n

'-

+ a

n0

P(l)
_>

By (11), we can find

k.

N so that when

Apply lemma 1 with A(x) = F(x),

oo
+ ^ x + . . . + V^1
+ X)
((1+
l + c)X
j xxx ^ 1 .
k=N+l

where, by Corollary 3,
In C ^ t - r in n
n
ln4'

Then apply lemma 1 with

A(x) = F(x), b(x) = ^

k=N+l

( l - ) X k x k _ 1

2(m-l),n

1966

AN ALMOST LINEAR RECURRENCE

This gives us

<p > C' where

n n

127

1 - 2
In C ~* (T /l)ln n
n
In 4'
Therefore
In 0

,
1

n
(In n)

is arbitrarily small when n is large enough.

Of course, the estimate we have derived in this theorem is verycrude as far as the actual value of

<t>
is concerned.
Empirical
n
^
tests based on the exact values of 4> for n < 10000 reveal exceln
~"
lent agreement with the following formula:
(20)

In < & J2-2L ( l n n . 2(l n In n) + 1) + In n - .843

n
In 4

The e r r o r is less than .05 for

n > 10; it reaches a low of about

- o 05

when n is near 50, then increases to approximately . 032 when n is

near 5000, and it slowly decreases after that.

Thus we can use (20)

to calculate
log 2 n
(21)

le
s 472n

12l

(1lIL)
in n'

with an e r r o r of at most 5% when 10 < n < 10000.

Although formula

(20) gives very good accuracy, it should be remembered that only the
first term of the expansion has been verified, and the comparatively
small values of ln ln n for the range of n considered makes it possible that (20) is not the true asymptotic result.

On the assumption

that the true formula is a relatively "simple" one, however, equation

(20) gives striking agreement.

A similar situation exists in the study

of the partition function; the methods used here can be applied with
ease to that problem, to give

log p(n) *v 7r\/-^

the actual asymptotic formula for

p(n) itself is

128

April

AN ALMOST LINEAR R E C U R R E N C E

W! n -I6
P(n) - (

4VT

j)

4wVr(n-^)

4V3n

rV

(n-l4)
where

p(n) <-

+ 0(e

A < ^ "|/

It i s d o u b t f u l t h a t it w o u l d h a v e b e e n g u e s s e d e m p i r i c a l l y i n e i t h e r of
t h e s e formSo

F o r a n a c c o u n t of t h i s a n d a b i b l i o g r a p h y ,

see f l l

REFERENCES
1.

Go H. H a r d y , R a m a n u j a n ,

2o

K, K n o p p a n d I. S c h u r , E L e m e n t a r e r B e w e i s e i n i g e r

c h a p t e r 8.

Chelsea,

asymptotischer F o r m e l n der additiven

Math,

N . Y. , 1 9 5 9 .

Zahlentheorie.

Z e i t s c h r i f t 24 ( 1 9 2 5 - 2 6 ) 5 5 9 - 5 7 4 .

xxxxxxxxxxxxxxx

NOTICE TO A L L SUBSCRIBERS! i I
P l e a s e n o t i f y t h e M a n a g i n g E d i t o r A T O N C E of a n y a d d r e s s c h a n g e .
T h e P o s t Office D e p a r t m e n t , r a t h e r t h a n f o r w a r d i n g m a g a z i n e s m a i l e d
third class,

s e n d s t h e m d i r e c t l y to t h e d e a d - l e t t e r o f f i c e .

Unless the

a d d r e s s e e s p e c i f i c a l l y r e q u e s t s the F i b o n a c c i Q u a r t e r l y be f o r w a r d e d
at first class

r a t e s to the new a d d r e s s ,

h e w i l l not r e c e i v e i t .

w i l l u s u a l l y c o s t a b o u t 30 c e n t s f o r f i r s t - c l a s s p o s t a g e . )

(This

If p o s s i b l e ,

p l e a s e notify us A T L E A S T T H R E E W E E K S P R I O R to p u b l i c a t i o n d a t e s :
F e b r u a r y 15, A p r i l 15, O c t o b e r 15, and D e c e m b e r 15.

SOME DETERMINANTS CONTAINING POWERS OF FIBONACCI NUMBERS

L. Carlitz
Duke U n i v e r s i t y , Durham, N.C.

1.

F . D. P a r k e r ( P r o b l e m H-46, this Q u a r t e r l y ,

Vol. 2 (1964),

p . 303) h a s proposed the evaluation of the d e t e r m i n a n t

'Fn+i+j-z'

(i, J = 1, 2, 3, 4, 5,)

This suggests the m o r e g e n e r a l p r o b l e m of evaluating

D, = D.
= |Fk, , |
k
k, n
n+r + s '

(r , s = 0, 1, . . . , k)
'

We shall show that

Ik(k+l)(n+l) ^
D k = (-1) Z
TJ

(1)

(*) ( F * F ^

. . . Fk)Z

j=0
F o r example
D ^ - t - l ) 1 * 1 , D2 = ( - D n + 1 2 , D 3 = 3 6

D = 13824,
4
To prove (1) we c o n s i d e r the q u a d r a t i c form
k
Q =

V* Fk , , u u .
smaj
n+r + s r s
r, s = 0

Since
F

a11 - p n

<a =

irrf-

1 + /5 "

2 '

PR =

1 - -/5

we have
k

(a_f3) Q =

uu us

k
k
( {_ 1 ) ^ "j / \a n (n+r + s)jp ( n - r - s ) ( k - j )

ZJ
Z^ r^ s 2^
JLJ
r, s=0

j= 0
129

130

SOME DETERMINANTS CONTAINING POWERS

April

=y\i) k -^Ya n v< k - j > E *(r+s)j p(r+s)(k-jVs

k

/ k

=^(-l) k -J(^) anJ pn(k-J)

j=0

VrJ

\r=0

If we put

r=0
it is c l e a r that

(3)

(a-^Q^^'fy^^^v)
j=0

Thus by m e a n s of the l i n e a r t r a n s f o r m a t i o n (2), we have reduced Q t o

diagonal f o r m .

If A denotes the d e t e r m i n a n t of the l i n e a r t r a n s f o r -

m a t i o n (2), it follows from (3) that

I

(4)

z^^1)r k
-k(k+n
ynk(k+i) 2
D k = (-1)Z
TT q) (a-P) k ( k + i ) (ap)Z
AZ
j=0

Now
A=

s
lc
I
|(arpK"r) I

(r, s=0, l , . . - , k )

Since this is a V a n d e r m o n d e d e t e r m i n a n t we get

A

A=

J7

, s,-1jbc-*s

0 <Lr < s <_ k

(a p

r_K."~r.

-a p

a r pr k " S (%
a - p )' F s - r
0 <. r < s <_ k
k-1 k - r
T7
(a_p)|k(k+l)TT
r=0

s=0

rpk-r-sF

!966

OF FIBONACCI NUMBERS
Ik(k+1) k-1
Z

= (a-(3)

JT a

r(k-r)

131
k

I(k-r)(k-r-l)
2

(3

"r

J!

r=0

s=l

Ik(k+1)
*k(k+l)(k-l)
= (a-P)2
(a-(5) 6

,
l

'

T h e r e f o r e (4) b e c o m e s
Ik(k+l)(n+l)
D, = (-1)

k
k
k_x
T T (*) ( F f F k
j=0

. . . F, )

This c o m p l e t e s the proof of (1)

2.

As for the d e t e r m i n a n t
Dk(L)= | L n + r + J ( r ,

B = 0 , 1, . . . ; k)

c o n s i d e r a t i o n of the q u a d r a t i c form
k
L , , u u
n+r+s r s
r, s = 0

X)^

aiij n(k j)

p " *(r+s)j p(r+

j=0

S)

<k"j>u

r,s=0

j=0
j=o

\r=0

u
r

yields
k

Dk(;D=rr anj

n(k j)

- A2

j=0
k
1
nk(k+1)
= (-l)2
J T ( k ) (a-(3) k < k+1 > ( F k F ^

...

FJ2

132

SOME DETERMINANTS CONTAINING POWERS

April

It f o l l o w s t h a t
i-nk(k+l) y k ( k + l ) k
,
,
, 1
Dk(L).= (-1)Z
5Z
T T (*) ( F * F * " 1 . . .

(5)

F^r

j=0
3

F o r m u l a s ( l ) a n d ( 5 ) c a n be g e n e r a l i z e d i n a n o b v i o u s w a y .

s i d e r the sequence

-J W J d e f i n e d by

W ,. = p W - q W
n+1
n
n-1
where

Wn,

vV,

Con-

are assigned.

(n > 1)

Put

Dk(W) = k | j

+ r + s

- 4<1

( r , s = 0 , 1, . . . .

k)

2
If

- 4q / 0

and

a = P + ^/P
then

= A a 11 + B p n ,

P - /p

where

W x - (3W0

A=

aW Q - W x

. ^

, B=

- p

'

AB =

fP

+ q W

^"7^

We find t h a t

Ik(k+1) Ink(k+l)4k(k+l)(k-l)
JL ,
Dk(W)=(-l)Z
qZ
T T (*)

(6)

'

(W

o yk(k+l)

1 "P

,
U

l ^V

,
U

<1 2

***V

'

where
U

- pa
a - (3

I n d e e d (6) h o l d s a l s o w h e n
U

= n(p/2)

- 4q = 0,

p r o v i d e d we now

take

T h i s c a n be p r o v e d d i r e c t l y i n t h e f o l l o w i n g w a y .

We

have

n = (f)n (A + Bn) '

1966

OF FIBONACCI NUMBERS

w h e r e A = W , pB = 2W

- pW .

D,(W) = |() < n + r + s > k

= (p)nk(k+l)+k

133

Then

(A + B(n+r+s)) k | (r, s=0, 1, .

(k+l) |

( A

k)

k|

B(n+r+s))

We r e c a l l that for a d e t e r m i n a n t of the type

|u , , |
i

(r, s = 0, 1, . e . , k)

+r+si

we have

k n + r +s l = I A r

uJ

(r, s = 0, 1,

OOBJ

k)

w h e r e A is the usual finite difference operator* (See for example

r
k
| 1, p. 1 0 3 ] , ) In the p r e s e n t i n s t a n c e u = (A+Bn) , so that
A1"

"
It follows that

=0
n

(r + s > k),
'

= ki B
n

|(A+B(n+r+s))k| = (-1 ^ ^

V l )k+1 Bk(k+1)

and t h e r e f o r e

(7) Dk(W) = (-ir ( k + 1 ) (f) n k ( k + 1 ) + k ( k + 1 ) ( k - 1 ) (k]) k + 1 (W 1 - f W0)k<k+1>

On the other hand (6) b e c o m e s

1
k
k(k+1)
n
.
l
)
2
P
)
k
(
k
+
l
)
+
|
k
(
k
+
l
)
(
k
l
)
j-j
(
(w> = < - i r
( P p i ^ i ^ M ^ T i , ^ - ! , -pj- (( -k}
j=0

Mw1-fw/<k+1>nJ2<k-J+1)(f)Z(J"1,(k"J+1)
j=l

(-1)

k(k+l)

(f)

nk(k+l)+k(k+l)(k-l)
TT (k) j 2 ( k " j + 1 )
j=l

(Wx - | W 0 ) k ( k + 1 )

134

SOME DETERMINANTS CONTAINING POWERS

April

Since

T7 <*> j2(k"J+1> =1^T <> ( - ^ y r ) 2 ( k " j + 1 ) = H ( > )

y

j=l

j=l

- I i
j=0

j=0

(k_j):

- I*-)

i t . i s c l e a r that (6) and (7) a r e in a g r e e m e n t .

REFERENCE
1.

G. Kowalewski, D e t e r m i n a n t e n t h e o r i e , Chelsea, New York,

1948.
xxxxxxxxxxxxxxx

All s u b s c r i p t i o n c o r r e s p o n d e n c e should be a d d r e s s e d to B r o t h e r U.
Alfred, St. M a r y ' s College, Calif.

All checks ($4.00 p e r y e a r ) should

be m a d e out to the F i b o n a c c i A s s o c i a t i o n or the F i b o n a c c i Q u a r t e r l y .

M a n u s c r i p t s intended for publication in the Q u a r t e r l y should be sent
to V e r n e r E . Hoggatt, Jr , M a t h e m a t i c s D e p a r t m e n t , San J o s e State
College, San J o s e ,
spaced,

All m a n u s c r i p t s should be typed, double-

Drawings should be m a d e the s a m e size a s they will a p p e a r

in the Q u a r t e r l y ,
bond p a p e r .
editors.

Calif.

and should be done in India ink on e i t h e r vellum or

A u t h o r s should keep a copy of the m a n u s c r i p t sent to the

ON THE QUADRATIC CHARACTER OF THE FIBONACCI ROOT

Emma Lehmer
Berkeley, C a l i f .

Let

$ = (1 +iJ~5)/2 be a root of the q u a d r a t i c equation

x 2 -x - 1 = 0

The n - t h t e r m of the Fibonacci sequence can be given by

(1)

^
11

Hence for any p r i m e

(2)

i}

0U

p / 5 we can state the c r i t e r i o n :

= 0(mod p) if and only if

02nE(-l)n

(mod p) .

If we define = 1 in t e r m s of the Legendre symbol as

(3)

e = (-) = 5 E l l
(mod p)
p 4 5
P
^
then a s p e c i a l c a s e of Lucas 1 t h e o r e m 1 s t a t e s
(4)

'=0(modp)

while a s p e c i a l c a s e of a t h e o r e m of L e h m e r
(5)

= 0(mod p) if and only if

2 gives
p=4m+l

Both (4) and (5) follow i m m e d i a t e l y from the c r i t e r i o n (2) and the
e a s i l y verifiable c o n g r u e n c e
(6)

"8 =

(mod p)

It is the p u r p o s e of this note to give a c r i t e r i o n for the q u a d r a t i c

c h a r a c t e r of

B and to apply it to find the condition for the divisibility

of
F

p-e
4

by p .
135

136

ON THE QUADRATIC CHARACTER

April

In the first place, if p divides

F

p-e"
4

then it must also divide

F

p-e
2
and therefore by (5) we have p = 4m+l, but since (p - 6)/4 must be
an integer,

= 1,

so that p = 20m+l,9.

The quadratic character of

6 for such primes is contained in the following lemma.

Lemma
if p = 10m+l

e ^~
(7)

Proof.
while
(8)

| -P-l

if p = 10m-l

Let a be a primitive fifth root of unity so that

a is an. integer modulo p if and only if p = 10m+l

= 1,

a / 1, then it is well known that

It is also clear that we can write 0 = - (a + a

a

) so that

a + 1 = 0 and hence

(9,

. - "2i>E

Considering (9) as a congruence modulo p and remembering that

is an integer for the primes under consideration we see that a will be
2
an integer modulo p only when Q -4 is a quadratic residue. But

e z - 4 - (e-2)(0+2) - - ( 0 - i ) 2 e^5

(io)
Hence

a is an integer modulo p if and only if

residue of p.
(11)

.
5 is a quadratic

Hence by (8) we obtain

I - v -

( ^ )

J 1

if

P = 10m + l

-1 if p = 10m-l

from which (7) and the lemma follow at once,

1966

OF THE FIBONACCI ROOT

137

But the q u a d r a t i c c h a r a c t e r oi\/~is

the s a m e as the q u a r t i c
c h a r a c t e r of 5 and this has been e x p r e s s e d in t e r m s of the q u a d r a t i c
partition
p = a2 + b2 ,

(12)

a E l ( m o d 4)

as follows |~3] :
^ i.s a q u a r t i c r e s i d u e of p = 4m+1 if and only if b = 0 (mod 5)
^ i? q u a d r a t i c ,

but not a q u a r t i c r e s i d u e of p is and only if

a = 0 (mod 5).
Hence our l e m m a leads to the following t h e o r e m .
2
2
T h e o r e m 1. Let p = a + b with a =_ 1 (mod 4), then
p-1
( 1 _if p=20m+l, b=0(mod 5) or p = 20m+9, a = 0(mod5)
v

p;

1-1 i i p=20m + l, a50(mod 5) or p=20m+9, b=0(mod5)

Combining T h e o r e m 1 with condition (2) for n = ( p - l ) / 4 we obtain

2
2
T h e o r e m 2. Let p = a + b with a _= 1 (mod 4), then
F , 5 0(mod p)
if a n d only if
~4~
e i t h e r p = 40m + l, 29

and b = 0(mod 5)

or p = 40m+9, 21

and a = 0(mod 5)

The p r i m e s

p ^ 1000

satisfying t h e o r e m 2 a r e

p = 6 1 , 89/109, 149, 269, 389, 401, 421, 521, 661, 701, 7 6 1 , 7 6 9 , 8 0 9 , 821, 829
The p r i m e s

p < 1000 for which

0 is a q u a d r a t i c r e s i d u e a r e

p = 29, 89, 101, 181, 229, 349, 401, 461, 509, 521, 541, 709, 761, 769, 809, 941
0 = 6 , 10, 23, 14, 82, 144, 112, 22, 122, 100, 173, 171, 92, 339, 343, 228
F o r some of t h e s e p r i m e s the following t h e o r e m h o l d s .
T h e o r e m 3.
a prime

If

0 is a q u a d r a t i c ,

but not a higher power r e s i d u e of

p = 10nl, then all the q u a d r a t i c r e s i d u e s of p can be gen-

e r a t e d by addition a_s follows:

r Q = 1, r x = @ (mod p), r^+l

= r^ + r ^ ^

(mod p)

138

ON THE QUADRATIC CHARACTER

This follows at once from the fact that

For example for

p = 29,

+$

April

=0(0+1)

= 6, all the quadratic residues are

1, 6, 7, 13, 20, 4, 24, 28, 23, 22, 16, 9, 25, 5

after which the sequence repeats modulo p.
Further results along the lines of Theorems 1 and 2 are as follows.
Marguerite Dunton conjectured and the author proved for p=30m+l
that 0

is a cubic residue of p, and hence that p divides

Vi '
3
if and only if p is represented by the form
p = s 2 + 135 tZ
The proof uses cyclotomic numbers of order 15 and is too long to give
here 0
709,

Such primes < 1000

811, 829 and 919.

are 139, 151, 199, 331, 541, 619, 661,

The author

4 has shown that

8 is a quintic

residue of p and hence that

is divisible by p if and only if p is an "artiad" of Lloyd Tanner C5j .

The artiads < 1000 are 211, 281, 421, 461, 521, 691, 881 and 991.

REFERENCES
1.

E. Lucas, Amer, Jn. of Math. 1, 1878, pp. 184-239, 289-321.

2.

D. H. Lehmer, Annals of Math. 1930, v. 31, pp. 419-448.

3.

Emma Lehmer, Matematika, v. 5, 1958, pp. 20-29.

4.

Emma Lehmer,

to appear in Jn. of Math. Analysis and Appli-

cations.
5.

Lloyd Tanner, London Mat. Soc. Proc. 18 (1886-7), pp. 214-234.

also v. 24, 1892-3, pp. 223-262.

xxxxxxxxxxxxxxx

ON EVALUATING CERTAIN COEFFICIENTS

Carolyn C. Styles
San Diego Mesa College, San Diego, C a l i f o r n i a

The coefficients

to be d i s c u s s e d a r e t h o s e i n v o l v e d w h e n e x -

p r e s s i n g t h e g e n e r a l t e r m of c e r t a i n s e q u e n c e s , d e f i n e d b y d i f f e r e n c e
e q u a t i o n s , i n t e r m s of t h e r o o t s of t h e r e l a t e d c h a r a c t e r i s t i c e q u a t i o n .
C a s e I:
71 .
(1)

m
anx

If t h e c h a r a c t e r i s t i c

equation

.
m-1 .
m-2
+a,x
+ a?x
+

+ a

^
m-1

+ a

= 0, a_ = 1
0

h a s no m u l t i p l e r o o t then

u
n
where

C, x
k k

= 0, 1, 2,
n

x. , k = 1, 2, . 8 . , m , i s a r o o t of ( 1 ) . If t h e b o u n d a r y c o n d i -

t i o n s a r e g i v e n by

(2)

= /
smj
k=l

u n = u 1 = . . = u m - 1, = 1
X

1X2

2 2
1X2

m m
l X2

k-i
2
X
k-1

m
X-i

l x

k+r "
2
X
k+1"

-L X -

k-1

lx2

2 2
1X2

k-1

m
2
X
m

1
i

then

m
, i e s a

k+1
X

k X k + l

2
2 2
k - 1 X k Xk+1

N
D

' *

m
2
m

m m
m
m m
X
l x2 '
k-1 Xk xk+l
Expanding the d e t e r m i n a n t s and dividing c o m m o n f a c t o r s
the n u m e r a t o r and d e n o m i n a t o r gives
x

139

from

140

ON EVALUATING CERTAIN COEFFICIENTS

April

m
1

(3)

N = (-I) ""

II

(x. - 1)

i=l

D= ( - l ) ^ 1 xk

(4)

(x. - x k )

i=l
i/k
Since

f(x) = 2^

a.x 111 " 1 =

i=0

II

(x - x t ) , a Q = 1

i=l

f(l) = O

(1 - x.) and f(xk) = n

i=l

(x fc - x.)

i=l
i/k

Using t h e s e i d e n t i t i e s , (3) b e c o m e s
N N

(-Dm+k^)
(1 - x k )

if
'

x
X

4 1
k

and (4) can be w r i t t e n

D=

(-l)m+kxkf'(xk)

Substituting t h e s e in (2) gives

Ck=l,
U

xk=l

xk(l - xk)f'(xk)

>

k^

P a r k e r [ 4 ] investigated the g e n e r a l t e r m of a r e c u r s i v e sequence

and gives a method for d e t e r m i n i n g t h e s e coefficients but does not give
the g e n e r a l f o r m u l a .
F o r the Fibonacci sequence the c h a r a c t e r i s t i c equation is
x

- x - 1= 0

and

u n = u, = 1

1966

ON EVALUATING CERTAIN COEFFICIENTS

141

Therefore
c

=
x

(-1H-1)
k^Xk " ^ ^ k " ^

=
2 x

i
k~l

Some characteristic equations obtained in generalizations of the

Fibonacci sequence and the values of C,

for each follow,

The characteristic equation in the generalization by Dickinson

[ l ] is x

- x

X f

k '^Xk^

- 1 = 0, a, c integers.
= CX

k"

aX

C,

k = ^ X k

^ '

Since

ax

*C " a ^ X k

+ C

'

=
(x k - 1) [(c - a)x k + c]

for the sequence in which u , = u, = . . . = u , = 1.

0
1
c-1
In the generalization by Harris and Styles [2J the characteristic equation is x^(x - 1) - 1 = 0, p, q integers, p 21 0, q 1 and
U
0=U1 = ' =Up+q-l = U
-

(P + q ) * k _ p

as was shown in j 2] without this formula.

Miles [3] used the characteristic equation
x

-x

k-1

- . . . - x - 1 = 0,

k integral > 2

For the sequence in which the initial conditions are given by

u

= u

k
~l
,
k
2x - (k + 1)
J

= u

k-i

= l

'

j = 1, 2, . . . , k

Raab [5~[ used the characteristic equation

r+1
x

r
- ax

- b = 0, a, b real,

r integral > 1 .

For the sequence in whichthe initial conditions are given by

uQ = U l = . . . = u r = 1

142

ON EVALUATING CERTAIN COEFFICIENTS

c

April

b +a - 1

=
k

(a - l ) x ^ + 1 + b [ ( r + l ) x k - r ]

The boundary conditions can be g e n e r a l i z e d slightly.

u Q = pr, u L = p r , u 2 = pr , . . . , u n = pr
C

C a s e II:

k '

If

n+1

x^(l - x k )f<(x k )

If the c h a r a c t e r i s t i c equation (1) has a root of m u l t i -

plicity 2 then
u n = (C x + 2 C 2 ) x J + 1 + ] T

C k x J + 1 , n = 0, 1, 2, . . .

k=3
and x~ i s the r e p e a t e d root of (1). If the boundary conditions a r e given
by3 u^ = u, = . . . = u
, = 1, then,
0
1
m-1
1-

(5)

x3

..

2x2

2
x3

..

3x2

3
x3

..

1l
x

2
2
x
2
3
x
2

"
m
x

1 2

m
2

mx

m
x3
x3

x3

3x2

3
x3

m
mx2

3
xm

'
*
..

m
2
x
m

2x

m
x3

m
x
m
x

m
2 !
x
m
3
X

m !

m
X

1966

ON EVALUATING CERTAIN COEFFICIENTS

143

Expanding the d e t e r m i n a n t s gives

m

(6)

m-1

N = n x
1

n (x - i) n (x - x .)
i

i=2

m-2

i
i=3

i=3

(
(x4-x3)

m i

n(x
i=3

..x.u.n (x.-x.)
m-1

i=3

(m-2)(m-3)
2

(-1)

(the sum of all p o s s i b l e f a c t o r s

( 2 x 2 - l ) n (x.-x 2 ) - x 2 ( x 2 - l )
i=3
(x. - x ? ) 5 i = 3, 4, . . . , m,

taken

m - 3 at a t i m e ] f
Since

f M (x 2 ) = 2 0

(x 2 - x ) and f , M (x ) = 6

i=3
(the sum of all p o s s i b l e p r o d u c t s of the f a c t o r s

(x

- x.), i = 3, 4,

. . , m, taken m - 3 at a t i m e ) , the quantity in the b r a c e s in (6) can be

expressed
(m-l)(m-2)
Z

(-1)

[(2x2-l)f'f(x2)

+x2(x2-l)f'"(x2)

Therefore,
(m-l)(m-2) m
NI

= (-D

11

i=2

m-1

x. 0
i

(x.-l)
i

i=3

II
i=3

(x

-x.)
m

m-2

n ( x m _ r x i ) - - - n (x 5 -x.). (x 4 -x 3 )
i=3

(2x2-l)f"(x2)

i=3

+ x2(x2-l) f',,(x2)

Expanding the d e t e r m i n a n t in the denominator gives

144

(8)

ON EVALUATING CERTAIN COEFFICIENTS

m

i=2

i=3

i=4

April

D = x 2 n x. n (x.-x 2 ) 2 n (x.-x3) n (x.-x4)

(x.-x
i

i=5

n (x.-x )

I
m-1
i=m-l
i=m
Substituting (7) and (8) in (5) and simplifying gives

m-2)

(i - x.)

2x2-l)f(x2)

i=3

C,

x2(x2-l)f'"(x2)

(x2"xi)2

i=3
If x 2 = 1, Cl = 1.

If x

4(i-x 7 r n

/l,

(i-x.)
1

i=3
1

(_2x2-l)f"(x2)

x2(x2-l)f,"(x2l

Lf'lx,)]^-^

Therefore,
J

2f(l]
x,tx,-l)f"(x,)

To d e t e r m i n e

i
1
1
"^x2)
1
1 T

[ x ? +t^ T ' r3+ f " ( x L )

2 ^

C ? the n u m e r a t o r in (5) is r e p l a c e d by

x2

2
m

m
x0

.
1

m
x3

= Nn

1966

ON EVALUATING CERTAIN COEFFICIENTS

145

Evaluating gives
m

(9)

N 2 = (-I)

n x. n (i - x.) n (x r x 2 ) n ( x r x 3 ) . .
i=2

i=2

(x.-x
i

i=3

9)

n (x.-x .)

i
i=m-l
i=m
Dividing (9) by (8) and simplifying gives

m-2

(-1) 1 " n (i-x.)

c

i=2

2
2

(x

i" X 2 V

m-1

2 n
=

m
x

i=4

(i-x.)

2
2-,,. '
x
2 "X2^

i=3
For

= 1, C = 0.

For

^ 1,

2f(l)
>2

To d e t e r m i n e

2fM(x2)(1"X2)
C, , k = 3, 4,
k

m,

the n u m e r a t o r in (5) is

r e p l a c e d by
X

2x
3x

m
x~

Evaluating

N,

yields

mx0

m m
x0

k-1
2
X
k-1
3
X
k-1

k-l

k+1
2
x
k+l
3
x
k+l

k+l

m
2
m
3

= N,

146

ON EVALUATING CERTAIN COEFFICIENTS

m

(11)

N, = ( - l ) ^

m-1

k-1
x

rt

-x.)

3
x

X )ee9

n ( k -i-2" i) P < k + l " i

i=2
i=2

(l-x2)

m-2

x? W x. T' (1-x.) "-M (x

i=3
i=2
i=2
i^k
i^k
i^k

k+1

April

p (x
i=2
i=k

.-x.)...

M (x 4 -x.) |II ( x 3 - x . )
i=2
i=2

(x,-x2)

i=3
Substituting (11) and (8) in (10) leads to
m
( - D ^ i - x ^
C,k =~

n (l-x.)
i=3

iA

T^T"
x, (x. - x 0 ) H (x, -x.) in- %(x.-x. )
k k 2 <
k i
i k
i=2
i=k+l-

(-l)k+m(l-x2)2 f
Ck =

-__-!_

(1-x.)
_

xkd-xk) -p (xk-x.)(-i) "

>

i=2
There

m k

(xk-x.)

i=k+l

C k = 1, x k = 1,
A1?,..
k-xk(l-xk)f*(^
If the r o o t s of
C
C

Summary:

if x

lf

4 I
k^L

m-1

f(x) = J ] a.x*" * = 0
i=0
a r e not r e p e a t e d and u = u. = u
^
0
1 2

= ,,. = u

. = 1,
m-1

xk / l

1966

ON EVALUATING CERTAIN COEFFICIENTS

x(l-xk)f'(xk)
If f(x) has a double root, x

xk= 1

'

k /I

= x , and u n = u
L

Li

= 1,
m-1
.

x2=l

f"!(x 2 )

2f(l)

x^xTTTF^)

J_ + _ L _ +
X--1

X?

2f(l)

^F7!

2 ^

= 1

fH

x^(l-x2)f"(x2)
X

C,

147

= 1

k = 3, 4, . . . , m

f(D

k ^

REFERENCES
Dickinson, David, "On Sums Involving Binomial Coefficients, "
A m e r i c a n M a t h e m a t i c a l Monthly, Vol. 57, 1950, pp. 8 2 - 8 6 .
H a r r i s , V. C. and Styles, C. C. , "A G e n e r a l i z a t i o n of F i b onacci n u m b e r s , " The F i b o n a c c i Q u a r t e r l y ,

Vol, 2, No. 4,

Dec. 1964 P
M i l e s , E. P . , J r 0 , " G e n e r a l i z e d Fibonacci Sequences, " A m e r ican M a t h e m a t i c a l Monthly, Vol. 67, I960, pp. 745-752.
P a r k e r , F r a n c i s D. , "On the G e n e r a l T e r m of a R e c u r s i v e Sequence, " The Fibonacci Q u a r t e r l y , Vol. 2, No. 1, F e b . 1964,

p. 67.
Raab, Joseph A . , "A Generalization of the Connection Between
the F i b o n a c c i Sequences and P a s c a l ' s T r i a n g l e , " The Fibonacci
Q u a r t e r l y , Vol. 1, 1963, pp. 2 1 - 3 1 .

xxxxxxxxxxxxxxx

, ADVANCED PROBLEMS AND SOLUTIONS

Edited by Verner E. Hoggatt, Jr.
San Jose State College, San Jose, C a l i f o r n i a

Send all c o m m u n i c a t i o n s c o n c e r n i n g Advanced P r o b l e m s and

Solutions to V e r n e r E. Hoggatt, J r . , M a t h e m a t i c s D e p a r t m e n t , San
J o s e State College, San J o s e , California.

This d e p a r t m e n t e s p e c i a l l y

w e l c o m e s p r o b l e m s believed to be new or extending old r e s u l t s .

Pro-

p o s e r s should submit solutions or other information that will a s s i s t

the e d i t o r .

To facilitate t h e i r c o n s i d e r a t i o n , solutions should be sub-

m i t t e d on s e p a r a t e signed s h e e t s within two months after publication

of the p r o b l e m s .
H-84

Proposed by Oswald Wyler, Carnegie Institute

F o r any positive i n t e g e r

of

Technology

m , t h e r e is an i n t e g e r

p r o p e r t y that m divides a F i b o n a c c i n u m b e r
divides n.
p-1

if

For a prime number

p = 1 (mod 5),

k(m) with the

if and only if k(m)

p, it is known that k(p)

and that

k(p)

divides

divides

p -f 1 if p = 2

(mod 5).
n u m b e r s , one finds that F,k(p)
. x
/ Usings a table of Fibonacci
2
is a multiple of p, but not of p , for s m a l l p r i m e n u m b e r s p, or in
2
other w o r d s that k(p) < k(p ) for s m a l l p r i m e n u m b e r s p. It does
not s e e m to be known w h e t h e r this is the c a s e for a i l p r i m e n u m b e r s
p or not.
H-85

Proposed by H.W. Gould, West Virginia University, Morgantown, West Va.

Let
D = f x n - ff x nJl
n
n
- n
y

where
F J_1 - f + f . with f = f = 1,
ii+1
n
n-1
0
1
and [~z"| = g r e a t e s t i n t e g e r
of z).

x = (1 + V5~)/2

< z (so that z - [ z ] = fractional

P r o v e (or disprove) the e x i s t e n c e of the l i m i t s

148

part

1966

ADVANCED P R O B L E M S AND SOLUTIONS

lim

149

D = 0 . 2 7 . . . = A
Zn

n > oo
and
lim
$> oo

with

A + B = 1.

G e n e r a l i z e t o c a s e of u

,, = p u

n+1
and u, are given.

and q are real and u n

H-86

D , ,, = 0 . 7 2 . . . = B
2n+l
+ qu

^ n

,, where

^ n-1

Proposed by Verner E. Hoggatt, Jr., San Jose State College, San Jose,

Let

p and q be i n t e g e r s s u c h that

Calif.

p + q _> 1, q >. 0

show that

if
x^(x-l)

- 1 = 0

(x-1)"
'

has roots

r,, r , . . . ,

and
x

= 0

n a s

roots

s , , s_, . . . , s ,
1 2
p+q

then
S

H-87

iP

q = r

i"

i = lf

2j

3, . . . , p + q

Proposed by Monte Boisen, Jr., San Jose State College, San Jose,

Calif.

Show t h a t , if
U

0
k

= U

= U

2 =

k-l

+ U

= U

'

k-2

= 1

n-l
+ U

---

k-n

a n d

k 2

'

then
1 - x

- 2x

7~
1

- . . . - (n-2)x
X

2
-

oo
> U Xk

=Z^ k
,

k=0
H-88

Proposed by Verner E. Hoggatt, Jr., San Jose State College, San Jose,

E
k=a

(n\

4mk V

T m

2k

2km

'

Calif.

150
where
bers,

H-80

A D V A N C E D P R O B L E M S A N D SOLUTIONS
F

and L
m
m
respectively.

a r e the

mth

F i b o n a c c i and m t h

April
Lucas num-

Proposed by J.A.H. Hunter, Toronto, Ontario, Canada and Max Rumney,

'London, England Corrected
Show

-0 *J
r=0

) F

2r+5

r =0

SOLUTIONS
THE F I N A L WORD
H-42

The c o r r e c t e d list i s :
1 , 2 , 3 , 5 , 9, 1 5 , 2 0 , 2 5 , 4 1

J.

D8 K o n h a u s e r f i r s t n o t e d t h e t y p i n g
Solution to the Crossword Puzzle

error.

by H.W. Gould, West Virginia

xxxxxxxxxxxxxxx

University

TERMINAL) DIGIT COINCIDENCES BETWEEN FIBONACCI NUMBERS AND THEIR INDICES

Gerard R. Deily
U.S. Department of Defense, Washington, D.C.

In [ l ] the Editor of this j o u r n a l proposed e s s e n t i a l l y the following question: What Fibonacci n u m b e r s of index l e s s than 10, 000 have
t e r m i n a l digits coincident with the index?

This note a n s w e r s that

question by supplying a c o m p u t e r - g e n e r a t e d table of such c o i n c i d e n c e s .

The c o m p u t e r p r o g r a m was w r i t t e n in the ALGOL 60 i n t e r n a tional a l g o r i t h m i c language [2 J and is given in the Appendix.

For

those r e a d e r s not f a m i l i a r with the ALGOL language, the o p e r a t i o n s

p e r f o r m e d by the p r o g r a m a r e b a s i c a l l y as follows:

With s t a r t i n g

values given, all Fibonacci n u m b e r s with indices 1 to 9999 a r e c o m puted modulo 10, 000. Fibonacci n u m b e r s

F,

through

F n a r e then

r e d u c e d mod 10 and c o m p a r e d with t h e i r r e s p e c t i v e i n d i c e s ,

through F q q a r e r e d u c e d mod 100 and likewise c o m p a r e d ,
i l a r r e d u c t i o n s and c o m p a r i s o n s a r e p e r f o r m e d for
FqqQ

and for

F, n n o

through

FQQQQ.

F, n n

a.nd

F,

sim-

through

If the c o m p a r i s o n yields an

a n s w e r YES, the index for which this o c c u r s is m a r k e d with an a s terisk

A 100 x 100 table ?

shown listed as tables l a n d II, is then.

p r i n t e d out with row c o o r d i n a t e s in h u n d r e d s and column c o o r d i n a t e s

in u n i t s , and with a s t e r i s k s in the locations w h e r e coincidences o c c u r .
Hence,

an a s t e r i s k in row z e r o column 61 i n d i c a t e s that

F,,

has

61 a s its last two d i g i t s , and s i m i l a r l y an a s t e r i s k in row 4 column 9S

i n d i c a t e s that

F 4 R ( - h a s 485 a s i t s l a s t t h r e e d i g i t s . Note the r e g u -

l a r i t y of p a t t e r n s in Tables I and II; t h e s e might prove to be an int e r e s t i n g subject for further i n v e s t i g a t i o n

A digest of the r e s u l t s r e p o r t e d h e r e i n is given in Table III.
These r e s u l t s w e r e checked through

F ^ Q C by inspecting a table [ 3 ]

of F i b o n a c c i n u m b e r s and by running s e v e r a l v e r s i o n s of the b a s i c

program^
REFERENCES
1..

B r o t h e r U. Alfred, " E x p l o r i n g Fibonacci N u m b e r s with a Calculator, " F i b o n a c c i Q u a r t e r l y , Vol. 2, No. 2 (April 1964), p. 138.
151

152
2.

TERMINAL DIGIT COINCIDENCES BETWEEN

Po Naur et al,

April

"Revised R e p o r t on the A l g o r i t h m i c

Language

ALGOL 60", C o m m u n i c a t i o n s of the ACM, Vol. 6, No. 1 (Janu a r y 1963), pp. 1-17.
3.

S0 L. Basin and V. E. Hoggatt, J r . , "The F i r s t 571 Fibonacci

Numbers",

Recreational Mathematics,

pp. 19-30. \&?i^t/iLe*6' <r>^

No. 11 (October 1962),

/S3^)

/O&MS

(Continued from page 116.)

U

k"

b = u

k-lUk-2---

V(k-l) "

k-(k-l)

Hence
i=k -1
u
Choose k~ < k 1

= (uQ - b)

II
i=0

u. + b

without loss of g e n e r a l i t y and the conclusion is a p -

parent.
We now c o n s i d e r equation (1) in s e v e r a l s p e c i a l c a s e s .

If b = 0

the equation is e a s i l y solved but in this c a s e the t h e o r e m holds t r i v ivially.

Let b = 2.

Then we have
u

k+i=

k -

2u

k+

which can be w r i t t e n in the form

k+i

- l - K -l?

The solution of this equation is c l e a r l y

(2)

uk = A

2k

+ 1

Hence the sequence with e l e m e n t s A^

.
+ 1, for i n t e g r a l A, is r e l a t i v e l y

p r i m e except possibly for the c o m m o n d i v i s o r 2

viously r e m o v e d when A is an even i n t e g e r .
F e r m a t n u m b e r s mentioned previously,,
(Continued on page 164. )

The exception is ob-

When A = 2, we have the

1966

FIBONACCI NUMBERS AND THEIR INDICES

APPENDIX
ALGOL-60 P r o g r a m for P r o d u c i n g Fibonacci T e r m i n a l
Digit Concidences

begin
p r o c e d u r e h o l l e r i t h (name, r e p r e s e n t a t i o n ) ;
real name;
string representation;
name: = representation;
i n t e g e r p r o c e d u r e mod (x, m);
value x, m;
i n t e g e r x, m;
mod: = x - m X (x f m);
i n t e g e r i, j , k;
r e a l blank,

star;

integer a r r a y

F[-l:9999] ;

real a r r a y cell[0:9999] ;
h o l l e r i t h (blank, ');
hollerith

(star, '*');

F[-l]:=l;
F [ 0 ] : = 0;
cell[0] : = star;
for k: = 1 step 1 until 4 do
begin
integer

finish, modulus;

modulus: = 10 ^ k;
finish: = modulus - 1;
for

i: = modulus / 10 step 1 until finish

do

begin
F[i]

: - mod ( F [ i - l ] + F [ i - 2 ] , 10000);

cell["il : = if mod ( F [ i ] , modulus) = i

then

end

star

e l s e blank;
end calculations for k - t h o r d e r of magnitude;
end setup of coincidence table;
w r i t e (for i: = 0 step 1 until 99 do
for j : - 0 step 1 until 99 do cellflOOXi + j " | );
F i b o n a c c i t e r m i n a l digit coincidence p r o g r a m ;

153

154

TERMINAL DIGIT COINCIDENCES BETWEEN

TABLE I
3Q

40

5Q

April

PAGE 5 OF 7
60

70

80

90
0
1
2
3
4
5
6
7
8
9
10

0
i
2
3
4
5
6
7
B
9
10
11
12
13
14
15
16
17
18
19
'20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50

12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
4Q
41
42
43
44
45
46
47
48
49
50

3Q'

40

5Q

6Q

7Q

SO

?0

1966

FIBONACCI NUMBERS AND THEIR INDICES

155

PAGE 6 DF 7
50

70

BO

90
50
31
52
53
54
55
56
57
58
59
60
&1
&2
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
S2
83
84
85
B6
87
88
39
?0
91
92
93
94
95
96
97
98
99

50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
8.1
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99

50

50

70

90

156

TERMINAL DIGIT COINCIDENCES BETWEEN

April

TABLE III
INDICES COINCIDENT WITH FIBONACCI NUMBER TERMINAL
1,

DIGITS

5,
25, 29,

41,
49,
61, 65,
85, 89
101, 125, 145, 149,
245, 265,
365, 385,
485, 505,
601, 605, 625, 649,
701, 725, 745, 749,
845, 865,
965, 985,
1105,

1205, 1249, 1345,

1445, 1585, 1685., 1825, 1925, 2065, 21

2305,

2405, 2501, 2545,

2645, 2785, 2885, 3025, 3125, 3265, 33

3505,

3605,

3745, 3749, 3845, 3985, 4085, 4225, 4325, 4465, 45

4705,

4805,

4945,

5905, 6001, 6005,

5045, 5185, 5285, 5425, 5525, 5665, 57

6145,

6245, 6385, 6485, 6625, 6725, 6865, 69

7105,

7205, 7249, 7345,

7445, 7585, 7685, 7825, 7925, 8065, 81

8305,

8405, 8501, 8545,

8645, 8785, 8885, 9025, 9125, 9265, 93

9505,

9605,

9745, 9749, 9845, 9985, others exceed 10000.

xxxxxxxxxxxxxxx

Also solved by Douglas Lind and Donald Howells.

PHI, THE GOLDEN RATIO (to 4599 Decimal Places), AND FIBONACCI NUMBERS
Murray Berg
Oakland, C a l i f .

The golden r a t i o can be obtained by the division of a line s e g m e n t into two p a r t s such that the r a t i o of the total line to the l a r g e r
s e g m e n t is the s a m e as the l a r g e r s e g m e n t to the s m a l l e r s e g m e n t .
This

r a t i o was given the n a m e ,

phi,

by the U. S.

mathematician,

M a r k B a r r , over 50 y e a r s ago [l~j

If B w e r e the l a r g e r s e g m e n t and A the s m a l l e r segment, then
A + B B
B
"A

(1)

Letting A = unity, we have

1 + B

~w~ -B
and the positive root, B is the n u m b e r , phi:

If we let B equal unity,

_ 1 +JE
z

then the r e s u l t would be one l e s s than phi.

P h i is the only n u m b e r whose r e c i p r o c a l is obtained by s u b t r a c t i n g

one from itself.
Phi is an i r r a t i o n a l n u m b e r , and its d e t e r m i n a t i o n depends on
the calculation oft/5 .

It can a l s o be calculated from the F i b o n a c c i

s e r i e s [ l ] and [ 2 ] .
The Fibonacci n u m b e r s a r e an infinite s e r i e s , w h e r e each e l e m e n t in the s e r i e s is the sum of the p r e c e d i n g two e l e m e n t s , and the
f i r s t two a r e equal to 1.
Thus
and for

Tl = F 2 = 1
i > 2

F. = F. . + F. . ,
i

where

i-l

i-2

F . = the ith n u m b e r in the F i b o n a c c i s e r i e s .

I

157

158

PHI, THE GOLDEN RATIO (TO 4599 DECIMAL

April

A p r o g r a m w a s w r i t t e n for the IBM 1401 which produced F i b onacci n u m b e r s up to 4000 digits in length. A p a i r of consecutive n u m b e r s w e r e p r i n t e d out at e a c h 100 digit level.
F . 7 / i s the f i r s t 100 digit Fibonacci n u m b e r .

F o r each s u c c e d -

ing 100 digit level, t h e F n u m b e r i n c r e a s e d by 478 and 479 a l t e r n a t e l y ,

i. e. , F Q . , is the f i r s t F i b o n a c c i n u m b e r with 200 d i g i t s ,
300,

h a s 400, up to F

with 4000 digits 0

F , .~~

nas

Table 1 l i s t s the

F i b o n a c c i n u m b e r s F , l n n o and E, i n n 4
It is evident from equation (1) that phi i s the r a t i o between two
consecutive m e m b e r s of the F i b o n a c c i s e r i e s .

T h e o r e t i c a l l y , we can

obtain the r a t i o to any n u m b e r of d e c i m a l p l a c e s by p e r f o r m i n g one d i vision with Fibonacci n u m b e r s that a r e far out in the s e r i e s .

How-

e v e r , this m u s t have p r a c t i c a l l i m i t s .
A s e p a r a t e p r o g r a m p e r f o r m e d such a division for F ,
F , , n n / t > both 2300 digits in length.
minutes0
versed.

inno

and

The division i n s t r u c t i o n took 21

This was r e p e a t e d with the n u m e r a t o r and d e n o m i n a t o r r e The r a t i o s coincided to 4598 d e c i m a l p l a c e s .

E m p i r i c a l o b s e r v a t i o n of m e m b e r s e a r l y in the s e r i e s show that

the t r u e r a t i o lies s o m e w h e r e between the r a t i o s obtained by the two
d i v i s i o n s . In fact, it is a p p r o x i m a t e l y one fourth the d i s t a n c e between
the f i r s t two non-coinciding digits of the two d i v i s i o n s .

These two

digits for the f i r s t division w e r e 42 and for the second division, 10.
The t r u e f i r s t digit of the two would be 3.
phi.

This i s the 4599th digit of

Table 2 l i s t s the fractional digits of phi.

REFERENCES

1.

M a r t i n G a r d n e r , M a t h e m a t i c a l G a m e s , Scientific A m e r i c a n , Vol.
201,. No. 2, Aug* 1959, pp. 128-134.

2.

No Vorob'ev, F i b o n a c c i N u m b e r s , T r a n s l a t e d from the R u s s i a n

by Halina M o s s , B l a i s d e l l Publishing Co. , 1961.
XXXXXXXXXXXXXXX

1966

PLACES), AND FIBONACCI NUMBERS

en

^
co
o

NO

OO

en
4svO
4so

OJ
sD
v>

en
Ln

o
oo
Oo
vO
00

11'

vO

t'

oo

co

-0
OJ

co
NO

-~J
00
00
4ss-

oo
1

4-

-J
00
OJ
0J
-0
OJ

ON

ON

-J
O

OJ

t'

Ln
O^s

ON

vD
Ln

tNJ

NO
H-'

t'

OO

-0

vO
00

ON

4-

4^-

NJD

00
OO
Ln

4^-

sO

ON
t'

-J

NO

Ln

NO

NO

tNJ
vO

H-

Ln
OO
Ln
O

00

ON

NO

Oo
Oo

co
o
Ln
CO
vD
IN)
O0
tNJ
OO

4s-vj

oo
o
SO

co

-0

ON
NO

O0
NO

Ln
H-*

Ln
s

00
00

4-

Ln
00
Ln
Oo

Oo

NO

co

co
4s-

Oo
Ln

vO

N)

00

1'

00
OJ

i"

Ln
H-

*""

NO

Ln
Ln

00

i'

4-J

o
o

OO
Ln
v>
OO
s

4s-

O0
-0
s

4-

Oo
vO

O0

vO
Ln

4*

4stNJ

-J
-J
-0
00
00
CO
s

4co

^J
i"

Ln

00

1'

4^

tNJ

o
H-

-J

en
0J

4s-

00

oo

to
-0
-0

1<

Ln
O

tNJ
tNJ

tNJ

4so

ON
ON

4-

I1

ON

o
o

OJ
Ln
CO

^ ]

h1

vO
tNJ
tNJ
-J
00

H-1
H-

-v]

4s-

Ln

,_.

tNJ
tNJ

-J

ON

oo

tNJ
v>
tNJ
tNJ

ON

00
Ln

tNJ

Ln
-J

H-.

00
Ln
Ln
Ln
vO
Ln

Oo
00

tNJ
-~J

NO

ON

tNJ

oo
4s-

Ln
00
00

o
o

tNJ

4s-

4-

- j

t*

-~o
o

tNJ

4so

1-4

o
00

ON
tNJ

-J
Ln

i'

os
4-

OO
Oo

OO

00

ON

ON

oo

ON

NO

ON

ON

os
4-

OO
00

00
Ln

H-

tNJ

OJ
tNJ

tNJ
[NJ

OJ

oo
o^ O N
o^ cr^

vO

4ON

00
Ln
Oo

00
-0
Ln

OJ

NO

4^-

4s-

vO

tNj
1'

NO

00
4sLn
vO
4s00
00
s

Ln
O

tNJ

-J
Ln
ON

O0
OO

,J

oo

H - '

Ln
Ln
fNJ

^J

NO

-J
O
OO
Ln
00
00
00
t-,

ON

H-J

-v]

vO
00

4-

4-

00
Oo
Ln
tNJ

ON

H-1

OO
tNJ

00

4^-

ro

oo

ON

OO

o-o
4^

oo
Ln
4s-

OJ
s

00
OO

- j

00

oo

4-

oo
oo

vO

~J

-vj

o
o^

oo

Ln
O

H-<

-J

o
4sLn

00

H->

I'

NO
h'

tNJ

4oo

Ln
sO

ON
CNJ

00
oo

-vl

Oo
sO
sO
s

ON

\>
4-

tNJ
O0

oo

1>

ON

1'

o
o

OO
-J

fNJ

ON
1

ON

Ln
-J

o
00

Ln

i'

4*-

oo

Ln
-J

tNj

-J

00

4s-

i'

00
s

o
Ln

4s-

o^

4s-

oo

H-'

ON

vO
00
OJ
vO

OJ
vO
00

1-1 4 -

-vj

fNJ

4-

00
00

oo
OO
-J
00
oo
t'

o
<l

400

oo

OO
vO
oo
-J
00

oo
Ln

vj
ON

00

N>

*-

O0

o
1*

ON

CO

I 1

o
o
oo

oo

OJ
O0
sO
4sLn
s

OJ

OJ

oo
00

00

vO
vO

00
ON

oo

-0
-J
tNJ
ON

o
Ln
o

oo

CO

ON
h-

Ln
H - '

<I
OJ
O^
OJ
OJ

OJ

K-'

N>

Ln

Ln
OJ

1-1

ON
h-1

Ln
-J
LO

00

K- 1

Ln

o
I

4s-

4*-

44sOO

oo
4ss4co
ON
-v)

- j

OO
O
OJ

vO

-J

Ln

NO

oo
CO

OJ
CO
Ln
0J

CO
s

4-

vd

1-1

H-.

H-.

('
4^-

-J
CO
Oo
00
CO
00
00
CO
Ln
OJ

oo
OJ
00

ON

4^
ON
-vj

00
ON

Ln
CO
v>
-J
00
s
4
00
I 1

Ln

Ln

-J

i'

ON

O^

Ln
vO
M

Ln
OJ

OJ

4^

CO
OJ
ON

NO

0J
s

o
o

-o
^J

-J

00
Ln
O
Ln
OO
i
00
00
OO

H_

H-i

4s-

NO

_.
00

vO

NO

vO

oo

OJ

Ln
00

vO

ND

oo

-v]

- j
CO
OJ

4-

Ln
tNJ
00

I 1

ON

t\J
00
OJ

NO

4-

oo

LO
vX>
-0

OJ
0J
CO

4s-

00

ND

i1

t'

-J
vO
OJ

CO
CO
CO

o
00
0J

o
4s-

-vj

ON

OJ

H-1

ON

H - '

4s-

-J
00
OJ

Ln

1-

co

oo

0O

oo
1

O0
Oo

1'

4-

h-1

Co

o
00

4=00

'-'
oo
CO
Ln
O
CO

oo
CO
-J
00
CO
Ln
OJ
OJ
NO

OJ
00
-J

oo

OJ
vO

00
00
t'

O0
-J

oo

OJ
-J
-J

1-1

o
OJ
ON

-J
s>
-J

NO

4-

t'

4-

Oo

i-u

OJ

00

4so
CO
4*o
CO

t<

-J

o
00
00
Ln
Co
Ln
Ln
4sLn
O
OJ
OJ

CO

4-

os

Ln

00
-J
vO

vO
Ln
CO

1'

OJ

4ss40J
s

4co

CO
s)

O
Ln

4s-

vO

I 1

-J
vO

o
oo

vO
Oo

vO

en

--3
vO

Ln

~ j

OJ
CO
v)

vO

o
co
00

'
CO

4-

00

4s4soo

co

o 4sos co
4 - 4s00
co OO
vD
Ln
CO

o
CO
o
4sv>

o
Ln
-vl
'

-J
0J
O
-J
CO
CO

-os
4-

>*

Ln
00

OO
OO
vO
-vl

o
o
00
OJ
-J
vO
00
Co
00

H-J

OJ
-^3

00

OJ

t'

NO

oo

o
OO

OO
Ln
OJ
00
OJ
00

-^3
s

4co
os
4s4-

OJ
CO
Ln
-J
OJ

vO

O0
OJ

-s]

Ln
0J

-v3

t1

OJ
LO

CO

4-

O0
CO
Ln
00
OJ

OJ
Oo
OJ
-J

Ln
O
00

4s-

o
CO

vO

H-1

oo

Oo

Ln
-^3

OJ
Oo

OO
vO
CO

!-

OJ
Oo

4sco

CO
CO

OJ
CO
OJ

Co

00

NO
NO

oo

H - '

4-J
CO

co
vO
CO
4sOJ

4>00

-a

NO

v)
OJ
t'

OO
Ln

. H -

^sO

4s-

H - '

H-'

O
-J
OJ

os

4-

Nd

t'

0J

1-1

OJ
OJ
vO
00
O0

00
-J
0J
CO
vO
OJ
4sLn

CO
CO
Ln

4so
o
OO

-0
CO
OJ

-~0
00
00

CO
-0
vO
vO

o
I

'

I1

o
Co

-0
O

sO

\r->

Ln
4s-

Ln

^J
0J
CO
vO
0J
Ln

oo

oo

Ln
vO

ON

H - '

4s-

-0
Ln

i'

vO

4-

00
00

vO
CO

NO

00
s

4-

00

--J

t<

oo

1'

~J

-J
CO

4^co
OJ

1'

OO

NO

CO

Ln

i i

N>

OJ

oo
4s-

00

I>

CO
xO

OJ

oo
4>-

co

-J
t

Oo

NO

- j

o
oo
^ j

H-.

NO

co 4--

NO

00

-J
Oo
CO

CO

4s4s-

4s-

4-

NO
1'

OJ

-J

i'

ON

Oo
vO

<o K- *
co
^n
00

4s-

Ln
CO
-J
CO

N>

Ln

OJ

h'

NO

4s-

4sNO

00

-J
OJ
O0
00

tj

ON

OJ
OJ

4s-

i<

00

vO

K-.

vO

4>-

Ln
0J
CO
00

OJ
-0
OJ
OJ
co
OJ

H-

oo

CO
Oo
Ln
Ln
CO

H-1

co
4s-

-0
OJ
vO

00

o
OO

OJ

-J
OO
CO

-J
OJ
CO

OJ
-J

CO
vO
0J

4sLn

Ln
Ln
CO

H-1

o^

vO
-J
Ln

NO

- j

en

-0
vO
-J

1'

--J
N>
NO

i'

H_.

OO
00

00
Ln
CO

vO
vO
CO

OJ
Ln
CO

4^

OJ
Ln

OJ
ON
Ln

NO

NO

CO

K'

-J

00
OJ
Co

-o

00^
NO

00

ON

OJ

00

H-i

CO

Oo
Ln
4sLn
O

oo
4>00

Ln
O
-^J
Ln

oo

4-

-~J

OJ
OJ
s

vO
OJ

i<

-J
Ln
-J
s

H--

OO
O

ON

ON

N>

vO
OO
OJ
CO
00

OJ
vO
vO

vD
CO

4-

OJ

4^-

vO
CO

oo
4s4s- - Js
co L4n-

ON

Ln
tj
OJ
1'

t'

oo
oo

OJ
vO

Ln
00
O

00

OJ

oo

rf-

ON

Oo
4sOJ
-J
OJ
CO
-J
s

Ln

-^j

I-J

Ln
-J

ON

OJ
O0
00

|J

o OJ
co oo

-J

Oo
-J
OJ

00

OJ

s>
-J

4o

O " CO

N O

Ln
00

4-

Ln
Ln

N O

ON

i 1

OJ
00
Ln
O

ON

ON
tNJ
OO
Ln

ON

ON

Ln
-J

00

Ln
s

oo

-vl

tNJ

1'

ON

l>J

vO
OJ
Ln
00
Ln
CO

ON

4-

oo

OO

t'

1"

NO

ON

vO
vO

00
v-O
vO
tNJ
Ln
tNJ
s

4o

- j

00

-J
-0

Ln
O
00
-J
OJ
vO

NO

OJ

NO

1_J

H-J

H-i

4co

CO

-o

i'

OJ
00
00
-^3
s

ON

-J
-J
OO

Ln
tNJ

-vJ

H'

4^-

4s- o
co OJ
o

Ln
OO
4sO0
vO

CO

oo

(4 -

Ln
O
-0
-J

t-1

OJ

--J

Ln
s

ON

-J

4s-

NO

O
OJ
-0

oo
4s-

1'
1'
1
00

00

N>

NO

ON

NO

i 1

Ln
Ln
O

Oo
00

OO
4sOo.
CO
-0

o
-vj

vO

-0
OJ
-J

vO
sO

on

NO
ON

o^
^o
Ln

i'

NO

00

4s-

vO

H"

4s-

t'

OJ
OJ

Ln

*'

Ln

NO

4soo

o^-

NO

Ln
CO
tNJ

00
00
4s00

4s-

os
4-

00

ON

^J
00

4s-0

*-'
4-

OJ
00

o
o

NO

00
Ln

0J
Ln

CO
OJ

i-,

4o
OO

OJ

o
H-

00
Ln
OJ

CO
00
00
s

OJ
-vl
OJ
Oo
-J

Ln
O

ON

CO

Ln
tNJ
OO

ON

Oo

tNj

ON

4-

OO
00

00
00
s

00 00

tNJ

-J

4-

4-

NO

Ln
s
4
00
OJ
Ln
i*
O
00
OJ
OJ

00

00

1'

^J

4-

4s-

4*-

t'

~J.
-J

00

N>

ON

t'

4ss-

-vl

Ln
tNJ
tNJ

vO

00
Ln

-J

v)
O

-~J
^J
tNJ
O
ON
H-'
tNJ

oo

-J
Ln
-0

Oo
O

tNJ

CO
-J
OO
OJ

o
tNJ

ON

IN)
OO
00

4-

00
4s00
-J

o^

tNJ
NO
-J
OO

ON

oo
~ j

oo

4o
os

t'

oo
o
Ln

-a
ii

00

NO

ON

xO
-J
tNJ
s

00
Ln

o
00

o
00

rNj

4^-

Ln

-^j

tNJ
-J

ON

NO

Ln
-J
Ln

-J

00

4so

Ln

-J

N O

00
-J
1'

4s00

4o
tNJ

v>

4s4soo

ON

ON

tNJ

Ln
4sLn

Ln
tNJ
Ln
-J
00
Ln
Ln
-J
-0
00

-v]

N>

vO
vO
00
tNJ
00
Ln
-J
tNJ

O
00
OJ

tNJ
tNJ
00
00

H-

ON

O
s
4
tNJ

1'

*o

OJ
vO

tNJ

Oo
Oo

NO

ON

oo
oo

o
o

H-"

N)

00

NO

tNJ

-v3

4>-

os

ON

oo
4sOO

4-

os

OJ

00
*
i'

OJ
O
00
00
Oo
Ln
Ln

4^-

Ln

1'

-J

ON

00

NO

ON

Ln
Ln

ON

v3
vO

ON

4o^
tNJ

-J
CO

ON

-J
00
Ln

,_J o^

Ln
O^

OJ
-0

4so

Ln
-J

vO
s

OO

-J

vO
Ln

OJ
--]

OO
Ln
tNJ
t'
00

o
00
s

4-

00

NO

vO
Ln
Ln
tNJ

ON
I'

-0

tNJ

ON

4s-

-0

ON

OO

o
t>

Ln
vO
Oo
Oo
sO
OO

OO
v>
sO

t>

-vl

t'

ON

ON

o
NO

ON

Ln

00

vO
tNJ
00

oo

4s-

4-

ON

Ln
-J

4-

vO

tNJ

t'

NO

00

-0

4^-

tNJ

00
00
s

vD
s

1'

00

ON

^J

ON

tNJ

sO
-J

ON

Ln
O0
Ln
-J

NO

-J
-J
O
O
~0
tNJ
00

4-

o
co

-J

NO

00
-J

tNJ
tNJ

00

I"

4s-

4s-

tNJ

4 - Oo
oo o
tNJ
tNJ
s

oo
oo

ON
vO
00

NO

ON
ON

ON
H-

NO

co
H-'
o

4s-

O
00
OO

00

-v]

fo
O
OO
4s-J
s

v>

Ln

Ln
Ln
00
^J

*_

0J
4sLn
00
vD
vJD

ON

tNJ
Oo
Ln
O

DJ
-J

00

4s-

tNJ

o
Co

t'

4o
-vj
H-1

os
4-

o ON
o^ o
tNJ

00

co
t'

tNJ

OJ
Ln
Oo
^-J
s

oo
o
v)

vO

^3

-J

OO
Oo

4s-

ON

o
00

H-'

H-.

00

NO

Ln

ON

Oo
tNJ
00

o
Ln

o
co
os
4-

u>
4sLn

H-

i'

Ln

N>
vO
OO

tNJ

vn

-v]

4-

NO

Ln

O
00
00
--J
00

Oo

Oo

H-'

OO
^J
00
tNJ
s

Oo
vO
Ln
O

OJ

ON

NO

NO

00

4s-

t'

~J
-J
CO
Oo
OO
tNJ
OO
sD

ON

NO

oo

sO

00

NO

-0
Ln
tNJ
tNJ
-J

4s-J

-vj
^3

4s-

oo

N)

159

ON

O^
H^

OJ
CO

o
00

4^-

co

Ln

Ln
Ln
OJ
OO
Ln
-J
vO

\'
4--

00
-J

o
00

H-

-J
OO

'
t'

4ss4^J
CO

'

o
00

'
4s-

sD
CO

o
Co

4
' OJ
- -J
s

oo

Co
CO
00
vO

oo
v
OJ
Ln
CO

4s-

CO
Ln

-J
1

~J
Ln
H-"

oo

NO

vO
vO

o
OO

160

PHI, THE GOLDEN RATIO (TO 4599 DECIMAL

o o- o
r- grD- 00
LO
CO
r-H
LD
00
o
co
r-H
rg r- o
Pti
CO rg LO
r-H
LO
g^
LO
o LOD
0 - rg
rLO
^
r- r-H a- rg
LO
tn 00 LO
rLD
co g D
LO
,
_
,
N>
r-H
CO
co
rr-H
fM
rg CO LO
CO
r- o v O
^
vO
rg g D
o
vO
rg LO
r- ^ ^
r(
LO
o 0o0 ^
r-H
0 O
o
0 '"3
O
co O rg
co
CO
ro
Io LO ^ rg
r-H
CO
0 r- LO
0 - r-H
co 0 - ^
r-H
LO
CM o
rg
o- LO g D r1 rg
co
r
co
o
co
r-H
gD
LO
LO
rg
0 - gD
co
r- o
rg r-H
rg o
IsCO
00 r-H v O
^f
I00 CO LO co
r-H
LO
rIH
^
00
o o ^
o
vO
rg rg r- 0 r-H
rg LO CO
rg
r-H
rg o
r^
0 O
co co r-H
o
r-H
sQ
co
^
sO
0 ILO
co
t-/* v D o
LO

^
Is

CO
00

r-H

vD
LD
00

rg
CO
O
Is-

LO
CO

O
LO
00

O
CO

CO

\F
o
LO

in

rg

eg

co
co
LO

CO

^
rg
rg

^
rg

o
CO

CO

o
LO

CO

o
LO

LO
gD

rg

CO

00

^
^
rg
^
CO

LO

gD

r-H

r-H

r-

LO

>-<

r-H

rg

eg
rg

co
r-H

Isrg

rg

0 -

rg
LO
rg

gD
LO

co
gD
sO
00
vD

00
00
00
CO
CO

r-H

r-

Xh

r-H

^
O

^
h
r-H
gD

rg
Is-

rg

o
rg

Th
CO
r-H

^
rg
rg

r-

o-

r-H

vO
00

rg
00
t-

o
o

00
LO
0s0s-'

o
r-H

CO

o
LO
o

r-H

r-H

0 LO
00
0s00

^
o

00

gD
r-H

CO

CO
gD

00

o
CO

00
gD

rg

Is-

Is-

gD

co
r-H
CO

co
o
LO
00
00

CTN

O
CO
gD

o
c
o
CO

r~-

r-H

C0
CO

LO

rg
r-H

LO

r-H

o
LO

CO

r-H

00

00

LO
CO
00

00

rg
LO
r-H

r-H

CO
o ^
co CO
r
rr-H
CO

CO
rg

l>

r-

O
O

r- co

r-H

gD

00
CO
LO
LO
00

LO

gD

o
o
00

r-H

r-H

o
rrg CO
gD
^H
o
rrg o
r- LO
o o

go
go
LO
0s-

gD
gD
LO

co
O '

Iss

co
o
go
o
Ir
Is

00

rg

r-H

^
CO
LO

o
o

gD
LO

00
LO

00

^
r-H

00

rg
gD

o o
r- 00
o LO
LO
oo rg
o

go

r-H

i>

Isr(

rI

r1

^
^
rg
r^
^ rLO

Is-

rg
rg
rg

LO

00
00

CO

,_4

00

LO

rg
rg
rg

gD

r-H

rg

^
^
CO
o ^
rg
^
^ ^
rg o
go
^
r- o
LO

^
o
rg

r-

sO
00
LO

o
o
rg
LO
gD
N)

00

00

r-H

oo

CO

LO

r-H

o
go
o
o
r-H

O
rg

LO

CT- o
00 go

00
LO

r-

0s-

r-H

r--

LO

o
o
ro
rrgD

r-H

00
0sLO

Is-

r-

go

^
^
rg
s

CO
LO
gD
CO

I-

i-H

0sCO

rg

^
^
LO

sO
00

xH

Is-

o
(D^

CO

CO
gD

Is-

CO

o
CO

rg

Is-

00

rg

rg

CO

o
o
CO

00
LO
CO

CO
CO
CO

rg
rg

0s-

LO
LO

CO

00
sO
LO

LO

rg

'-,

rg

o
o

LO

o
LO

0s-

rg

^H
^
_, _
rg
tgD

0sCO

O
Is-

rvO

r1

gD

LO

r-H

CO
gD
ri

00
r-H
s

I r1

oo o
gD
gD
rg

^
H
LO
Is-

o
o
LO
CO

go
00

00
rg
rg

CO
00
LO

^ ^
oo

gD
gD
LO

xh
00
0soO
00

oo
co
0 s

r-H

00

00

LO

r-

r-H

r-H

00
Is-

rg
rg

oo
r-H

rg

r-H

00
00

0s-

00
LO

CO

rg

0s-

co
sO

rg

Is-

rr-

0s-

IsLO

o
r00
r-H

LO

o
oo

gD

xO
CO

oo
CO
IsLO

LO
CO
vO

ro
0 s

LO
0svO

o
0 s

0sLO
0s00
0s-

LO
CO

rg

O
0s0s0s-

rg

00

^
^h

LO
gD

r-l

LO
Is-

o
rg

gD
LO

rg

rg

-JD
IsgD

o
o

LO

^
xH

vO
0sCO

rg
rg
rg

0s0s0s-

rc
o
rg
0s-

gD

r- o
ao
rg 0 0
r\

rg

CO

0s-

gD
LO
CO
0sLO

CO
00

^
f
00

r-H

rg
LO
CO

rg

r^
r- ^
o rg

o
oo a^
F
r-H
0s-

^
^H

CT-

00
LO

ro

0s-

rg
Is-

r-H

^
o

rg

rg

rg
rg
go

LO

oo
go

LO
LO

0sri

LO

rg

vO

Is-

LO

rco

LO

0s0sgD
0s-

LO

gD

0s-

t\

00

00

oo
r-

CO
00

^ rg
'-'

^
f r
- o rg
0 LO
co ^f
LO
go r- rr-H
xH o r1
rg 0 0 0 0 LO
s

CO
r-H

^
s

0 0s-

CO
00

0s00

gD

^
^

go
gD

o
0 rgD
s

LO

oo
00
LO
00
00
0s00

Is-

o
LO

,
_, I rg gD
s

^
H
r-H

LO
LO

LO
Is00

o
LO

LO
0s00

00

go

gD
0srI

rg
Isrg

rg
vO

rg

-^
0srI

r-

^ '
rg
^ gD

gD

Is-

CO

oo
o
CO
LO

rg

o
o

0sgD

LO

rg

co

^
O

go
gD

^
rg

0s-

gD

gD

go

rg

go

LO

gD

o
LO

ro
go

vO

00
00
LO

o 0o rco
co LO
i>
^
o 0rg0
00
s

gD

rg
LO

rg

-*

0sgD
rg

rg
r-H

LO

rt

oo

r-H

TH
LO

Is00
00

o
Is

LO
LO

^
o

r-<

00

00

00

gD

gD

rg

rco

i-H

0sgD

^
H
0 -

Is00
00
CO

Is-

LO

^H
r- ri-H
rrg xF
rg

gD

Io o
o 0 - 00
^F rg oo

CO

LO

r-H

o
^F i-H
O
o
o
o LO
LO
00
r
o- g D LO LO
CO 00
o oo
00

LO

^
^
rg rg

oo
r-

IsgD

--

rg
gD
oo

go
rg

00
0s-

O
r-H

^
o

Is0sIs-

-<

r-H

I -

00

CO

CO
rg

-*

r- rI
r- 0 0
o
oo
I - g D .o
oo I - ^H
0 o 00 LO LO o
gD
o 0 - gr-HD 0o- ^
LO
o r-H
oo
0 rI
00
CO ^
co r-H
r{
sO
CO
00
o rg
00
rg CO
LO
^
F
0 LO
LO
I-

CO

gD

0s-

r-H

,_, oo
rg

Is-

rg

o
rg

i-H

CO
0sCO

o
o
00

r-H

00

o
o
rg

^
oo

0s-

0sgD

Is-

MD

rg

rg

r-H

i>

LO
LO

r-H

00
0s-

00

LO
00

00

go

^>

00

rg

rg

rg

rg

rg

O
r-H

o
o
rg

r-H

LO
gD

rg

xf<

r-H

gD

LO

o
LO

00
LO

LO
LO
0O
sO
0s-

CO

o
a->
oo
rCO o
rg
o
^
co
r
co
r-H
00
0s-

rg

co
CO

co
o
oo oo
O ; oo
r-H
0 -

CO

Is0s-

r-l

gD
rg
go

0s-

r-H

r-H

rco
r-H

LO
Is00

gD

00

0s-

gD

^
^ "^ ^
o
^
CO

gD

0s-

CO
gD

00
0sgD

r- ^h

00
LO
0sLO
00

o o
oo 00
r-H
r^ ^

r-H

co

00
00

gD
CT^
CO

LO

co
o
LO
CO
LO
^
CO
oo co oo
o
r-H
0 gD
^
r-H
^
CO
Ir
- o
r-H
00
LO
O
o
gD
00
O
o CO
XF
o r{
o ^
^
r-H
LO
Is00

^
^

LO
0s-

rg
Isco
roo

Is-

r-H

CO

r-H

r-H

rg

rg

<
tf o
r-H
o
v>
rg

LO

LO
00

00
0sCO
00

0s-

CO
00

r-H

r-H

r-H

go
rg

,_, r> ,
_, co~
r-

r-H

Is0sgD

LO
,_< oo
oo ^
H rg rg
r- 0ri- gD CO
o
r-H
gD
o
r- I ILO oo rg rg
s
00
LO 0 - I - ^F
co
CO
o LO 00 rg CO
r-H
oo rg o r-H
xH
rg 00
r- rg o go
LO
rg rg CO
rg

rg

CO

r-H

LO

r-\

LO
LO
LO
CO
LO

00

^
^
o go
00
^ co
^

00
00
gD
LO
g^

r-H

r-H

Isr1

o
["o
rg

00

LO

co

00
Is-

00

co
o
^
CO
r- gID- LO
r-o o, o
o
CO
ILO
o
'- go 0 0 o LO
LO
LO
o
^
r-H
LO
o r-H 00
^
0 00
rg go
o
00
Io LO ^
o r-H
i-H
o CO rg go
xh ^
rg rg CO ^
rg
gD rg r- r- o
Isrg 00 00 rg
r^
0 gD
LO
\h LO CO
r- co CO o o g D
rg oo o
00
LO
o
LO
o rg
oo g D o
rg I - . o
co
o
^
r-H
r-H
gD
Io
LO
o g^
^D r-Hr-H or-Hr-H grg
rD- o
^
rg 00 O
rg

^ ^
^ orrg rg

r-H

gD
CO
00
00

0 0sgD

rg
Isco

00
gD

o
oo
LO

gD
00
gD

^
H rg
r- xF
CO
co
LO
r- rgo r- CO
go rg
rs
CT"I - rg
gD
o rLO
rg g D

o
o
crr-H

'-

o
o
rg
00

00
gD

0sgD

rg

gD
rg

ICO

LO

go

r-H
r
IsLO
o o
o rg
LO
co
o
0
rg
I -

gD

r-H

LO

^
,_,
^ rg

uO
gD

--" r-

CO

i 1

LO

o
CO

00

sJD
O

r-H

co
o

o
LO

0 r- ^
r-o go
o g^
r-H
D
00
o go
gD
LO

rg
F-

LO
gD

r-H

go

r-H

rg

oo
xp
o
rg

00
CJN

00

gD

gD

^
o

r-H

r-H

rg
rg

CO

CO

gD

xp

rsO

sO
LO

gD

CO
MD

gD

rg

o
o
o
00
o
LO

o
o
o
CO

rg

LO

o
ri-H

rg

r-H

CO

gD

co vi>O
cr^
o o
o o
00
LO

o
NO

co
o
I -

l>
LO

>sO

CO

Tt
LO

o
o
o

o
o

r-H
r-H

LO

xh

CD

go

r-H

rg
<{

LO
LO
LO

00

gD
rg
gD
CO
gD

rg

rg
rg

00

vjD

CO

o
o
LO

00

CO

rg

00

oo

rg

co r-H
o
r-rg O

go

LO

roo oo
o 0o0
rg o
O
a- r-H 0 0
rg LO ^
o
gD
o
^
LO
rg ^
rg r-H
r-H
g
D
co 0 0 LO 0o0
o
rg o
,_, rg
LO
LO

c
o
rg

00

r-H

CO

CO

rg

'O

CO

LO
VJD

^
o

rg
^
ro
^ ^
ro ^
LO
00

gD

o
o
o CO

^
CO
LO
^
o oo
o
rg g D
rg

^ -'
00

co

CO
LO
LO
LO

LO

arg

r- ^
o
rg

CO

gD
CO

LO

o
r-H
o
CO oo o
g
D
LO
ro 00
o
rg r-H
o
0 - LO
LO
rg o
O
LO

CO

rg

^ rg
^
rg ^
oo
rg

Ti

rg co
r- 0^
gD
0
r- o o

LO
CO

LO
CO
LO

00

rg

r-

'-

r-

r-H

00

gD

0s-

r-H

r-H

^
LO

l>

co aLO
oo
o CO
rg co
gD
oo
rg LO
r- o
o CO
r- o
^ rr- LO
r- o
o r-H
r-H
rg

o
o
rg r-H
oo ^ O CO
o grD- r-H
r- o
ov
o vc^
r- ^
O
CO
o
r- ^
o ^ rg ^
,
_, LO org rg
LO
o o sF
gD
00
o
r-H
LO
rg
o
LO
rg
r- rr-\
0 - CO
o
v-O
LO
r- sH r-H
r-H
rg
o
r-H
v
O
LO
0 - CO
, co

rg

LO
00
gD

co

0s00

00

^
xH

IsLO

o
gD

00
O

00

rg

LO

gD

o
o rg
o
r^
LO
co ^
r- o
co
co
CO
00
oo ^

o
xh

o
o
CO
rLO
O
rf
o
r- o
rg rg
r- rg
rg
rg

0s-

r^P s O
-tf co
o rgD rg

^
o
^
LD
rr- g D
o CO
o

vO

rg
00

April

rg
LO
0s00

^
r-H

rg

00

1966

PLACES), AND FIBONACCI NUMBERS

^
(
^

I1

^1

NO

NO

00

O
00

NO

vO

-vj

ON

O
00

On

4^

00

on

^]

Oo

4sOO
IN)

I'

INJ
O
00

On

-v]

H-"

IV

tv
OO

o
00

NO

OO

4^-

y'

00
NO

OO

t\)
00

IN)

< !

-J
00
40 s0y'

ON
I1

O^

o
o

o
oo

On
On

00
-J

NO
O

OO
IN)

ON

t'

00

NO

i1

00

oo
o
o

u>
tNJ
OO

4^
on
o
o

NO

Oo

4^

-J
00

1'

IN)
H-'

I1

INJ
vJD

oO N
4^
tV

NO

ON

00
^J
00
tNJ

00
00

(^
oo

< !

NO

4^- On

4^
o

ON

oOO

4N.
Ul

On

^ j

NO

|>

00
IV
IV

oH-

tv

On

On

On

H-

NO

00
00

Ui

oo

^
00
N)

4^-

~J
-J
00
-J

00

NO

On
On

00
00
O
-J
IN)
00

IN)

On

oo

Q N

-o
o

00
Oo
IN)
OO
--J
OO
f

Oo
O

-0

y>

NO

NO

^j-

tv

O0
00

On
On
IV
00

^J

IN)
y>

IV

Oo
[V

tv

IN)
O0
00

^ o-

o
4^Oo

H - '

4*-

NO

On
NO
l

On

NO

00
O
NO

oo
oo
o
NO
IN)
00
-J

On
^J
O
O0

Oo

oo

Oo

^
('

OO

NO

ON

NO

I1

hf^

00

-0
rN)

Oo
Oo

-J

NO

t'

oo

00
ON
00
O0

On

oo

INJ

NO

^
1<

OO
H-'

h^

-J
NO

00
On
-^J
i'

NO
I1
H- 1

OO
00

On

I4

On

OO

I1

NO
-0

O
Oo

Oo

1J

o
o^
00

OO
ON

00

IN)
O

^
On
^
On

H - '
1'

hl^
00

NO

On
o

OO

oo

NO
NO

Oo
00
On
On
On
-J
CT-

1^
NO
~0
-J
NO
tNJ

'

NO

oo

OO

IN)
< !

NO

oo

ON
IN)
ON

oo

oo

-a

^a
o^

-J
tN)
IN)

1-1

I-"

-J
On

-J
On

i*

NO

NO

NO

00

4^

00
ON

ON
tNJ
-0

On

IN)
NO

tNJ
H^

-o

I1

Oo
On

tNJ
-J

NO

On

tNJ

tNJ
ON'

^J
OO
00

On
On
On

-J

On
On
O

^J

On

4^

On
00
O0

On

ON
t>

00
-0

H^

OO

on

^J

4^
00
tNJ
ON

-J

"1'

00
1'

4^H-

4^

NO
00

oo
00
00
h'

IN)

o
o
o

ON

tNJ

Oo

oo

NO
t'

ON

IV

NO

NO

-0

4^

tNJ
ON

Oo

4^tN)

O
NO

Oo

o
oo
00
ON
-J

4^

IN>
NO

On

-v]

00
O

On

oo

4^

oo
ON

--J
00
-J
tNJ
O0
tNJ

On

On

H_

tNJ
OO

4^
4^4^
o
o
ON

-J
IN)
<1
-J
-J

4^

4^
o

IN)
O
OO

tNJ
ON
00

NO

Oo
On
On

4^

On

On

On

OO
ON

-0

H-4

Oo

oo

ON

ON

4^-

On
ON

IN)
NO

00
00

NO

tNJ

-<

00
001

I
00
ON
00
00

4^

4^
o
o
on
tNJ
tNJ
tNJ
t'

ON

NO

On
-J

NO

-0

ON

O0

I1

On

OJ
00
ON
ON

tNJ

4^

ON
00

O
ON

oo

tN)

o
on
tNJ

On
INJ
OO
O

NO

4^

-J
-J

On

NO

4^

otNJ

ON

OO
ON
tNJ

NO

N O

t>

rf^
OO

INJ
OO

-J
On

4^

-J

oo
-0

4^
IN)

00

NO
NO

4^-

NO

ON

On
On

IN)
-J
-J
00
ON
tNJ

ON

NO

On

^
^
on

00

ON

4^-

tNJ
t

tNJ
-J

ON
i'

NO

~J
O

ON

H-"

NO

tNJ

On

00

IN)

NO

4^

H-"

ON

on

on
ON

4^ON

on
00

IV
00

4NNO
NO

oo
o

tNJ
NO

00

oo

ON

NO

H-

tNJ
4NON

NO
NO

I1

00

00
tNJ

on

On

NO

I<

ON

~J

NO

on

NO

-J

-J

ON

H-"

oo

4^oo
ON

00

O0

4^

H-1

-J

oo
4^

oo
H'

On
IN)
O0

ON

oo

^J

o
on
^J

Oo
i

ON
ON
tNJ
On
ON

4^tNJ
-J

4^
o
~o

o
rf^ oo
-J

INJ
i

NO

On

00
NO

ro

O0
tNJ

^3
-J
v'

y1

00
NO

O
IV

4^4^

NO
-0

-vj

IV

-v]

tNJ

o
4^
y'

00
NO

oo

O
IN)
ON
-J

NJO

00
00

~J

oON

4^- ~J
4^ On
4^- tNJ
o
oo
4^- - 0

K-.

oo

NO
INJ

o
4^
4^

NO

On

ON

4^
oo
ON

4^-

NO

O0
00

OO

CNJ

tNJ
ON

NO
N O

tJ

-J
NO

^
4^

ON

INJ

tNJ
00

On

00
OO
IN)
tNJ
ON

oo
I

4^
00

4^
4^

-J

on

-vj

NO

On

t'

tNJ

'

ON

00
vO

On

O0

O
On
IN)

4^-

Oo

vO

t'

< !

ON

4^oo

IN)

On

tNJ
IN)

^J

~0
^J

oo

o0 0

On
00

NO

I1

On

on
00

4^o
o
-0

tNJ

OO

NO

4^

tNJ

o
o

y'

^J

00

t'

4^- oo
oo 0 0

NO

--J
ON
INJ
00
ON
IN)

t'

on

NO

On

ON
ON
-J
tNJ
ON
OO

--J

--J
00

--J
ON

ON

-J

On
*
4^

4^4^
oo

OO
O

~<i
NO
00
O

ON

-^]

NO

oo

ON
OO

00
O
O

oo

NO

4^
N O

00
ON
tNJ
ON
IN)
NO

o
NO

V'

< ]

-J

NO

4^-

>'

-J
-J
tNJ

H-.

o0 0

OO

OO

NO

oo
00

('

-J

NO

On

00

4^-

4^OO

on

ON
NO

ON

Oo

ON
ON
~J
00

4NNO
O
ON
00

t1

'-'

4^

OO

4^- On
O0
4^oo 4^
00
oo
00
ON

o
o

ON
-J

on

OO

tNJ

00
00

o
-0

NO

00
-<l

ON
NO

ON

<1

00

N O

tN)

On
Oo

-0
00

'-'

On

tJ

tNJ

tNJ
OO

o
on
ON

oo

On

ON
ON
NO
^J
NO
00

ON

-J

00
4N-

>'
4^

4NON

on
OO

NO

NO
NO

h-i

-0

00
-J
~J

On

--J
tNJ
-J

ON

4^
4^- On

4NON
ON
ON

tNJ

4^

OO
OO

NO

ON
IN)
1'

oo
oo

I1

NO
-J
ON
ON
OJ

O
NO

I1

4^

oo
o

tNJ
ON
NO
ON

4^
4^

tNJ

OO
ON

NO

1-1

-0
ON
-~J

O0

On

4^
4^

tNJ

^J
N)
OO

On
O
O

ON

tv

On

<l
ON
tNJ

IN)
IN)

ON

ON
ON
00
NO

IN)
On

O
H

--J
t'

'

00
ON
00
OO

tNJ

i1

OO

NO

oo
oo

NO

4^

On

oo

On
-0

00

NO
00
-J
4N-

to

Oo
IN)

H-'

Oo

IN)

-J

-J

1'

00
tNJ

y>

IN)
ON
ON
00
OO

4^- o
ON
4^
^a
00
00
00
On
O

On
tN)

ON
tNJ

I1

ON
tNJ

00

,_.

ON
NO

00

tNJ
00
NO

NO

NO

oo
-o

o
on

H'

oo

4N-

^3
00
4NOn

-^J
-J
<l
00

H-

NO

1'

t'

00
00

00
1'
ON
tNJ

4^-

on

NO

-0

ON
tNJ

-vj

~J
-J

IN)

4NINJ

^J

* 00Oo
(-- 4^
On

On
ON

-J

00
tN)

NO

On

I1

ON

NO

f^

o
4^

t-,

NO
NO

h - '

^J

--J

00
On

t>

o
o
oo
OO
00

tN)
00

NO

On
On
IN)
O
00
On
tN)

--J

1'
1'

On

t\J

ON

NO
ON

oo

-J
<l
tNJ

ON

o
on

On

~J
On

t-- 4

-J

NO

INJ
tN)

NO

f'

INJ
OO
-0

i1

00

ON

oIN)

On
On
00

tNJ

t'

H-'

-vl

ON
tNJ
IN)
ON

on

tNJ
IN)

4^on

On
On
IN)

DO
NO
4N.
ON
00
NO
ON

H-'

ON
tNJ
-J
NO

>-'

NO

NO

OO

-v]

t'

00

NO

OO

tN)
OO

h - '

4^

*-'
4^

O0

NO

h-^

tN)
O0
tNJ

o
o

NO

ON
-J

oo
4^

NO

On

tNJ
ON

ON
NO

o
o

tNJ
Oo

H-'

On

00

'

-0
00

NO

O0
00
NO
00

1'

-vj

NO

1-4

1^

tNJ
-J

1'

00
00

tNJ

on
t

ON

*' On^

On
O

IN)
tNJ
tNJ
OO

00

On
On
tNJ
tNJ
IN)

-J

-sj

H-

OO
<l
00

o
o
o

4N-

oo
-o

On

4^
4^

tNJ

OO

NO

On

ON
NO
^J

O
IN)
tNJ

4N-

4^

Oo
On

Oo

t'

4^

ON

IN)

On

00

i1
H--

ON

1'

NO
00
ON

~0

-0

tNJ
tNJ
tNJ
NO
00

00
O

4^
4^

4N.

IN)
H-1
00

On

On

ON
00

4^
o

,
Oo
>
oo

tNJ

tNJ
t^
00
tNJ

ON
ON

H-

i1

-J

On

K)^

00
ON

OJ

oo

1'

NO

ON

-^J

on
o

ON
tNJ
1'

4^

tNJ

4^-

ON

INJ
IN)

oo

4^4^

IN)
ON

NO

h'

i1

4^-

4^
4^-

Oo

NO

00
00

00
tN>
NO

-J

tNJ
tNJ

On

'-, o*
oo

ON

)>

On
O
-v]

NO

*' ^ 4^oo
^
^
o
-

IN)

o
o

~vJ

tN)
^
4*-

00

o
o

NO

NO

Oo

On

-v]

NO
ON
00

oo
o

NO
NO

00

ON

00
-J
ON

00
O0
ON

o^

-J
-J
tNJ

t ^

ON

oOO

OO

-0
On

NO

00
NO

on

oo

-J
#>>Oo
J^
_.
on ,on

i'

NO

On

tNJ

ON

oo

h'

o
o
4^
o
o

On

00

I1

OO

H1

H-.

-J
On

On
On

On
O

O
On
O

oI

NO
tNJ
00
00
-J
-0

'

00
00

On

o
oo
h->

NO

ON

*
H- 1

r\J

-vJ
00

Oo

-'

NO

NO

OO

NO

On
On

-vj

On

OO

0sOn
tN)

tNJ

o
rf^ On

On
NO

On

OO
NO
-J

NO

On

NO

oo

oo
4ss400

' ^
^
v^-

oo
o

NO

vO
^J

H-'

On

tNJ
00
IN)
tNJ

ON
IN)
O

Oo
00
-0

Oo

N>

On

ON

OO
O

On

4^-

On

-J

-vj

On

NO

4^

NO

OO
tNJ

ON
ON
-J

tNJ
4NINJ
NO
tNJ
ON

On

On

00

ON
-0

NO

4N-

O
00
00

00

NO
O

O
ON
-J
ON
00
NO

NO

4^

OJ
tNJ
O
00
IV
-J

NO

Oo

o^

00

On
O
OO
00

On

O
--J
<1
ON

00
NO

-J

ooo

00

oo
otNJ

ON

H-

f^

-J

1-1 l-J

On

On

00

00

-vj

tNJ
Oo

00

NO

1'

ON
h-1

Oo

NO

00
-^J

NO

1'

^J

NO

' oo
4>' <!

y1

00
On

o
oo

ON

IN)

IN)

O N

-J
-J
ON

y>

NO

00

ON
ON

hJ tv

h-

OO
o

00
ON

ON

On
IN)

i1

NO
IV

00
--J

H--

^J

IN)
NO
OO

*-

Oo

o
oo

tNJ

rf^
>o
4^

o . O
on

^
o

tNJ

00
NO

ON

4^- On
oo N O
-J

On
O
IN)

ON

^ ^

-v]

NO

On

h - '

NO

00

On
On

H-1
NO
00
00

OO
IN)

00

^
^
oo

NO

ON

oo
o

tN) NO
00 H^
00
Oo
IN)

IN)
INJ

ON

1>

OO
-0

-0
O
NO
00

,_,

On

-J

4^
o

On

ON

o^
00

o
On
O

--0

00

O
O^

NO

IN)

-v]

-'

ON

I1

Ul

H-.

i1

*
IN;
^ o^ (^
On o

o
ro
OO

v>

^J
00

i ^

4^
ro
o
4^

NO

iv

ON

00

-0

oo

--J

4s-

NO

oo

4N-

-J
vD

NO

NO

o
o

On
On

00
O
--J

O
0sOO

-J

O
00
00

t-

NO

NO
00
-J
00

On
Oo

On

NO

On
On
^ ^
^
H lJ
"~

o^

H-

OO

OO

-v]

IN>
IN)
tV

NO
ON

NO
NO
NO

NO

NO

o
oo
tv

Oo

00

-0

EN>
O

On

On

-J

O^

NO
NO

On

^
<!
O
O

y'

f^

NO
00

t'

on
I
1

On

00
tNJ
IN)

INJ
O
H^

*"* On

On

On

-vl

NO

NO

IN)
-J

00

t'

On
O

t'
1

4^
00

oo
o

NO

<I
00

Oo

(^

Oo

i'

OO
00

-v]

ON

N)

t'

Oi

IN)

ON

tv

NO

^ ]

On

^J

OO

^3

I-.

NO

-~j

-J

Oo

11

4^

4^
o

00
ON
NO

NO

*-'

NO

^J

ON

*->oo

O0

hf^
On

-vj
H-

IN)
-0
O0

H- 1

CNJ

-J
s

4n

' ioo

i'

-vj

INJ
OO
00
O

161

y'

y>

oo
..

& ^
On

O
oo
oo

^
^
oo

O
-J
On
00

00
-O
>I>
sO

oo

oo

oo

NO

^
oo

-J

00

ON

H -

vO

162

PHI, THE GOLDEN RATIO (TO 4599 DECIMAL

Is-

CO

r-H

vO

r-

v)

LO

LO
CM
r-H

ri

co

vO
CM

r-H

LO

rH

CM

^
o
^
^
vO

CM
CO

o
r-

sD

F1

-f

00

CM

CO

00

CM

CM

CM

Is-

-vf

rON

^
r-

00

CO

r-H

ro

o
r-H

it

o
CO

r-H

LO

LO

LO

00

00

#
^
CM

LO

ro
o

o
LO
00
CM

vJD

LO

CM

CO

r-

CO
CO
CO

CO

CM

sO

i>

vJD

Is-

CM

o
r~
o

a-o

CM

CO

r-H

o
r

roo

00

LO

CO

o
o

ao
o

O N

vO

CM
vD
00
O

r-H

00

TH

I -

r~

LO

ON

r-s

vO
CM

LO

CO

vO

r-H

CO
O

CO
ON
ON
ON

N O

v)

r-H

CO
Is-

LO
IN-

r-H

O
00
r-H

xh

LO
r-H

CM

NO

CO
sD
00
CM

NO

"tf
o
rr-

r-H

r-

vJD
r-H

LO

CO

CO

LO

CO

co

O N

CM
vD
LO

LO

I -

CO

CO

LO

CO

00

r-H

"*

CM
CO
IsCO

CM
CO
vO

CO
SO

LO
vO

CO
CO

LO

O N

LO

LO

LO

XH

vO

-*

LO

vO

CM

r^

LO
vO

CO
X}H
ON

CO

CO

^f

CO

CO

LO
CO

^ ^
o
LO
ON
CM

r-H

(M
CO

'
VJO

r~H

LO

rLO
ON

CO
LO

O N

CO
r-H

CO
ON

00
vO
CO

LO

r-H

CM

CO

_H

^
^

co
r-H

CO

LO

LO

LO

co

CM

LO
CO
ON

r-H

r-

00

r-H

LO

CM
CO

NO

vO
LO

r-

r-H

CO

00

IsCO

r-H

o
o

fM

,_,

LO

I -

r-H

QN

ON
r-H

( N -

UO

sO

O
00
l>
00

LO
CO

CM

CO

,_,

ON

vO
00

co

co

vO

CM
CO
ON

LO

l>
CM

LO

00
CO

MD

CO

CM

NO

LO

r-

NO

LO
sD

CO

r-

oo
s

00

LO

^ ^
r~
O

CO

CO
vO
ON

00

CO

00
CO

t-

co

00

LO

r-

LO

LO

00
r-H

Is-

^f
LO

CM

O
r-H

Is-

LO

o
LO

vO
vO

^
o

vO

CO
r-H

sO

CO

CO

CM

I >

^
r^-

00
CM

CM

CM
r-H

CM

ON

ON

^o
r-H
ON
ON

ON

CO

<sh

r-H

r~
o

r-H

LO
ON
r-H

CO

r-H
ON

00

sO

CO

r-H

CM

CM

00
r-H

CM

r-H

rco

r-

ON

r-

CM

ON

ON

LO

00
CO

LO

CO
vO
ON

oo
Tj<

00
r-H

r-H

CM

CM

ON

CM
LO

CM
MD

NO
rH

LO
CO

I -

r(M

CM
00
CO
Is-

ro

NO
CO

r-H

^ ^

CO

CM

vO

CO

sO

CO

xp

^
O
^
rvO
^
t>
CO

00

CO

N>

sO

ON

^
s

I 00
IsIs-

rLO

r-H
ON

CM

LO

LO

LO

CO

CO

r-H

LO

CM

Is00
O N

ON
LO

ON

O
vO

r-H

v>
sO
s

I MD
sO
CM

Is-

00

00

vf

CM

CM

Nf
ON
O
r-H

CM

CM

CM

CM
r-H

LO
CO

00
ON
CM

CO

r-H

O
CM

xH

XH

LO

CO

XH
r-H

^
o
"tf

LO
Tf

vO
XH
vO

Is-

CO

00

CO

CM
rt

oo

CO

ON

CM
r-H

r-H
ON

CM
Is-

co
r-

CO

vO

r-H

CM

CM
ON

co
(M

sO
CM

tf

ON

00

CM
CO

00
IsIs-

O
O

^ v^O
o

CO

CM

sD

00
ON

CO

CM

Xf

CM

LO
r-H

r-

CM

CM

ON

LO

(M

N O

Is-

XH

LO
r-H

Is-

(M

LO
CO

r-

r~i

CO
ON
ON

CM

CO

r-H

CM
r-H

CM

00

CO

vO

r^
(M

LO
ON

LO

CO

LO

vJD

LO
r-H

sO

00

Nf
LO
ON

^O

00
ON

o
CM

vO

Q N

r-

00

ON

v>
IsIs-

^
vO

-X)

r-H

vD

00

r-H

vO

CO
vO

r-H

sO

CM
v>

t-

Is-

CO

Is-

CO
CO

CM
l>
Is-

O N

LO

r^

Is-

^
CO

xH

CO

CO
ON

ON

r-H

CO

r-H

CM

r-H

CO

ON

Is-

CO
CO

IsI>

ON

CM

Nt<
ON

LO
sO

co
CO
ON

r-H

CO

CM
ON
LO

LO
ON

^
o

LO

00

co
o

CO

CO

CO

Is-

^
^
o
o
r-

LO
CM

LO
CM

CM

r-H

LO

LO

f-

r-H

00

CO

o
r-

CO

LO
r-H

o
00
CO

^ ^
XH
-<
^
sO

CO

00
r-H

NO

LO
^H

vO
IsIsCM

CO
CO

xf

v>
r-H

CO

00

fM

r-H

00

LO
ON

Is-

LO

"-'

CO

vO

LO

co

CO
vO

Nt*

CO
sO
r-H
ON

CM
CO

l>
CM

s>

LO

CM

vO

ON

Is-

r-H

00

cO

ON

co

LO

CO

CM

NO

t>-

MD

r-H

NO

ON

o
CO
CO

00
sO

ON
CO
CO

vO

CM

LO
MD

LO

LO

CM

CO

^
^
oa
o

MD

N>

CO
CM

CO

CO
CM

XH
CM

sO
O

LO
ON

-X)
XH

CM
ON

CO
vO

ON

CO

vO

CO

CO

co

r-H

ON

Is-

I -

o
o

Is-

sO

CO

00
ON

N O

xH

[N-

CM
r-H

ON

00
r-H

CO

CO

LO
r-H

LO
ON

LO

I r-H

00

CM

CO
CO

vO

CM
r-H

LO

r-

CM

LO

XH

[N-

Is-

CM

00
ON

CO
r-H

LO

00
sO

00

ON

LO
CM

LO
r-H

LO

r-H

CM

ON

ON

CM

00
CM

sO

LO

r-H

CO

r-H

sO

00
Is-

LO

r--

r-H

00
vO

I CM

CM
ON

CM

^
LO

CM

CO

NO

r1

LO
CO
ON

l>-

r-H

v>

cco

00

Is-

^H

00
r-H

CM
ON

vO
TjH

CM
r-H

CO

'^
-*

CM

I >
MD
r-H

r^

LO
vD

ON

r-H

CM

ON

LO

r-H

vO

00

00

co

-JD

T O

CM

CM

"sH

CO
ON

CM

CO

O
CO
ON

xh

co
LO

CO

^
r~

'-' ^ ^
^H
^
^
^ ^f
^
vO
^ ^
o

CM
r-H

CO
IN-

LO

CM

O
CM

CM

LO
CO
ON

t>
CM

ON

CM
r-H

N Q

LO
ON

LO
CM

00

r-H

vO
CM
00
00
00

r-H

LO

CO

r-H'

CO

ON

I >

sO

CM

ON

r-H

CO

CM

CO

vO

O N

r-H

vO

r-H

ON

00

00

[N-

LO

CO

r-H
ON

r-H

LO

ON

LO

XH
CM

LO
ON

ON

NO

Isr-H

sO

CM

vO

r-

CM

00
ON

LO

ON

LO

vO

xH

CM

LO

r-H.

ON

r-H
ON

r-

v>

r-

I ON

v>

r-

sO
CO

CM

CO

co

ON

ON

NO

if)

-<

co

LO

O
00
f>
CM

vD

LO

LO

CO
ON

r-H

ON

vJD

CM
Nt*

ON

Is-

r-

r-H

^f

CO

N O

Is-

CO

LO

00

vO

CO
ON

^
o

CO

CO

CO

ON

O
sO

^
^
o

XH

t>
^
^
^
o
^
^
r~
x^ "*
_, o o a )
.^
o
^
o
o
(M
I

^o

I 00

r-

co

r~
LO
CO

xj-f
CM

LO

r-H

CM

^H

O
CO

^ ^
v^
O
O
^
o
Nf
^
^
o
^f
00
^ co
^H
o
00
^
o
sO
CO
CM
CM
co
^
^ ^
co xh ^
rXH CM
o
co
^
vO
,_, LO
LO
o
co
fM
^ oo
^
o
o
o

CO

CO

LO

CO
r-H

CO

ri
CO
I>
Is-

r-

CM

r-H

o
co

CO

r-H

vJD

LO

ON

(M

[ >

CO
ON

CO

r-

xH

CM

os

LO

vO

LO

v^
O
^
o
o

vO
CM

CM

LO
r-H

(M

CO
LO

CM

00

Is-

LO
CO

00
r-H
ON

r-H

CO

r-H

LO

co

I -

LO

ON

o
o

r-H

vD

CO

00

o.

r-

CO

LO

LO

r-H

ON

CM

LO

CO

r-H

O
LO

[N-

co

LO
r-H

Is-

vO

ON

QN

LO

LO
vO

CO
MO
vO
CO
Is-

ON

LO
CO

CO
NO

NO

CO
r1

CM

r-

i-H

NO

00

sO

CO

I -

CM
r-H
CO

NO
CO

co

00

MD

I -

LO

CO
Is-

ON

I -

vO

LO

LO

CO

r1

CM
CO

T4H

XH
CO

CM
Is-

CO

LO

ON

r-H

LO

LO

co
'-<
o
,co_,
r-

r-H
ON

^
r-

Is-

Is-

LO

CO

CO

O
ON

XH
o

LO
ON

o
LO

CO

r-H

r-

r-H

CM

CM

CM

I Is-

r-H

sO

NO

^
O

CM

CM
CO

00
i>
CM

CO
r-H

LO

NO

00
r-H

CM

CM

Ttf

r-H

ON

co

xf

CO

LO
CO

v>

"^ ^
o

v>

r-

^
00

LO

O
LO

r-H
ON

LO

N Q

co

00

vO

CM
CO

LO

r-H

r-H

xh

CM

Nf

LO

00
Is-

CO

r-H

LO
ON

O
CO

LO

Isr-H

rr-

LO

CM
r-H

CO

CO

Is-

00

00

CM
|N-

XJO

CM

CO
r-H

s)

CM
Is-

s)
ON

O
ON

MD

t>r-H

ON

^
r^ co
o
^ ^
r-

LO
vO

LO

a--

ON

^
o
o

CO

CO

I -

IsON

LO
ON

CM

o
o
os

ON

CO

00

CO
ON

NO

LO
CM

^H

ON

r-

IssO

"Nf
ON

LO

r-H

rr-

IsON

LO
-HH

r-H

CO

CM

r-H

CM

CO

IssO

IsrH

vO

rc^-

O N

r-H

ON

(M
H
ON

IsON

00

v>

00

r-

r-H

'

vO
vO

r-H

ON

LO

r^

r-H
r-H

ro
o

O N

CO

ON

o
ro

vO

v>

LO

r-H
ON

-*
vO
^ ^
^ ^
^
o
o
co
^ co LO
co ^ vO
CO

00

ON

T*4

r-H

CO

r-

fM

CM

vO
00

r-

CM
r-H

CO

CM

CO

vO

LO
LO

CO
[N-

LO

CM

s>

vO

r-H

CO

r-H

oo

r-H
ON

i>
C0
LO

ro

CO

rr-H
ON

CO

ON

^
r-

( N -

r-H

r-H

Is-

CM

CM

CO

(M

CO
CM

CM

o
xH

r-H

os

r-H

LO
sO

CO

o
r4

r-H

vO
CO

ON

rx>

r-

^
CO

CO

o
o

NO

co

oo
o

^
^

^
o

CO

CM

O
ON

CO

CO

vO

r>
xH

CO

ON

r-H

r-H

CO

LO
r-H

OJ

LO

LO

^
^
o

LO

iH

ON

sO

CO

LO

ON

r-H

vJD

v>

vO

r-H

r-H
ON

ON

r-

0s-

LO
sO

sJD
ON

00
ON

Is-

CO

sO

r-H

v>

CM
CO

CM

CO

00

CO

LO

I -

r-H
ON

vD

CM

LO
CM
r-H

xH
rr-

ON

CO

CO

ON
00
NO

CO

r1

CM
Is-

r-

CM

N O

LO
ON

LO
LO

00

^
Th
r-

CM

CO

CM
vO

MD

^H

00

o
o
o

N D

CO

"*

^H
xH

CO

o
o
co

CO
CO

CO
vO

LO

CO

00

(M

O
rr-

00

LO

LO
iH

tN-

CO

LO

CO
r-H

00
LO

^
r-

00

l>

00

vO

ON

CM

CM

00
CO
O

LO

00

CM
r-H

00
[N-

r-H

CO
CO

LO

ON
T^

CO

CO

co

CO
r-H

xH

CM

ON

00

CO

CM

r-H

CM

ro
o
r-

co

00

CM

CO

CO
XH
vO

LO

LO
CO

LO

ON

CM

ro
CO

r-H
Isr-H

__, ^
^
CM
ON

r-H
ON

ON

00
vO
ON

ON

Is-

Is-

CM
O

r-H
r-H

CO
00
00
Is-

^
CO

00

Is-

00
ON

CM

ON

CM

CO

LO

LO
r-H

CO

ON

CO

LO

CO

CM

Is-

^o

LO
00

LO

LO

fM

CO

Is-

LO
vO

-*
Is-

LO

xh

CO
r-H
r-H

CM
00
vO

CO

xH

CO

MD
ON

sD
LO
CO
LO

NO
CO

ON

LO

CO
O

00
CO
sO

^
oo

co
co
o

CM

ON

CM

00
ON

Is00
00

CM
ON

LO

o
o
^o
xh
CO
r-H

00
r-H
LO
CM
LO

April

POWERS OF THE GOLDEN SECTION I

Robert S. Beard
Berkeley, Calif.

2
The Golden Section is the positive root k of the equation x + x = l ,
k = ( N/~5 - l ) / 2 = 0.618034.
The negative root,
k.

(- v/lf - l ) / 2 , is

- 1 / k , the negative r e c i p r o c a l of

F r o m the above equation,

,n
. n+1 . ., n+2
k = k + k =

,n-Z
k -

,n-l
,

That i s , any power of k is the sum of the next two higher p o w e r s or

the difference between the next two lower p o w e r s .
If the p o w e r s of

a r e listed s e r i a l l y as in the box tabulation

with the accompanying d i a g r a m , the ascending and descending r a t i o s

of the s u c c e s s i v e p o w e r s a r e

k and 1/k

respectively.

E a c h power of k can be e x p r e s s e d in t e r m s of its f i r s t power

and a Fibonacci n u m b e r a s indicated in the right column of the box
tabulation.

Starting with the s u c c e s s i v e p o w e r s

and k , this

column can be completed by r e p e a t e d application of the k

It is evident from this tabulation and the k

formula.

formula that the p o w e r s

of k form a Fibonacci s e r i e s which can be s e p a r a t e d into two c o m ponent Fibonacci s e r i e s .

Continuing f u r t h e r , a l l power of k can be e x p r e s s e d in t e r m s of
any other power and a n u m b e r a s shown on the accompanying t r a n s formation tables.

The box tabulation on the d i a g r a m can be used to

d e t e r m i n e the values given in the t r a n s f o r m a t i o n t a b l e .

the value

For example,

k = -5- k

- -5- is obtained from k = 5 - 8k. This value for

0
8
k, coupled with the value k = 0 + , can be used to d e t e r m i n e all of
the values listed in the v e r t i c a l k" column with the aid of the k
formulas.

It is interesting to note the recurrence of the Fibonacci sequence

in the n u m e r a t o r s in the v e r t i c a l columns and in the d e n o m i n a t o r s in
the h o r i z o n t a l c o l u m n s .

Both the Fibonacci and Lucas s e r i e s a p p e a r

163

164

POWERS OF THE GOLDEN SECTION

April

in the k v e r t i c a l column., Complex e x p r e s s i o n s involving v a r i o u s

p o w e r s of k can be v e r y m u c h simplified by r e f e r e n c e to t h e s e t a b l e s .

REFERENCE
R o b e r t S. B e a r d ,

"The Golden Section and F i b o n a c c i

Numbers",

S c r i p t a M a t h e m a t i c a , Vol. 16, M a r . - June, 1950 pp. 116-119.

XXXXXXXXXXXXXXX
.
(Continued from page 152.)

//

In g e n e r a l we m a y t r a n s f o r m equation (1). by writing

(3)

u k = F(V k )

Suppose that f u r t h e r m o r e we r e q u i r e that F satisfy the functional equation

F 2 ( s ) - bF(s) + b = F(2s)

(4)

Then our equation b e c o m e s

F(V

j^.

given by V
(5)

) = F(2V ), a solution of which is

K+l

= A2 .

Hence we have
u k = F(A2 k )

We now c o n s i d e r the functional equation (4).

F(s) - = 2L(s)
Then equation (4) b e c o m e s
(6)

L 2

<

h2

>=I

(Continued on page 169. )

- | ) + M2s)

Let

1966

POWERS OF THE GOLDEN SECTION

Hf

4-

-sn
1

i - | = H* Nlfi HS
Oi

X i

xr

4-

1 +~
0>

U1
In

4*

OJ

XT
X
1 " i+n *

XT
4-*>

in

IP

r|
X,

X.

N|Vfl

wl- til
I
+
r>|< N K

Oi

XI

Plf^

tin
+

tsi

XT
+ Os

00

In

X",

X,

X",

11
0>
4>

+r>

U>

I
NI09
X ^

HOi

XT
i "

4-

X",

HS HS -12

MS -Is

r>l(AJ

1
r*l-

Do

CM

rIIrf

ro

rJ|-j

x+ ~f

MS HS
XT

0>

r
1

-^

Cn

fi

H 1 * n.ir.

rlS

u+

-J

OD

CO

N|C

Xi

+ <p
I

XT
4-^
r*>|u

rlt]*l

IS!
+

1
n|*

|MfO>

ro
XT

XT

-1

f-

rl-

r.|-

t|ro

U.J

wl

4-

M0

rv)|-

0>

ro

r
XT

I
XT

x,

+ **

i "

1
N l -

NlM

H-

x ^ XT
1"

4-

o
f

XT
+ ~

X i

+ "
o

X,

XT

j K

4-1

Nl*

ro|i*J

o
+
Hr<s

l | -

x,
I

H-

X,
f
wl-

1 1

n>|ln

N[(P

tHl

4rH

rH^

U
XT
4r^

X i

x, ,i

4-

H=

4-

Oa

Ml

Cn

OS

X,

x, j

XT

ns

h|ln

rules
XT
4>
rlN

l *

N ^ 1N|0
X, I
4M|O

,!

XT
+

H-

raC0

^1

tsi
+

4-

0 I
XT
+ -

x~,

-l-

r*

X,

03

X * i

XT
+ "

4-

UJ

XT
I
in

X i

0> J tn
,

"

08

4*

X,

XT

I
t

1
|

At

HS HU!

| N

X".

X,

^ 0

Hs
(n

4 - -

FWensd K.

HE -IR
tsi ^1
+ +

HS -IS

^15 ^ I i o

4-

Cn

t
1
CD

1
CO

un ta

1*1-

XT

Nlm

PIOt>

n>

Oi

In

nJ

xr

+~

"*

rl-

165

+H 4r|b Nlin

XT
4-
H<

x-,
4 -

^il

HS

pa
f

HS

HS

X i

+ *

U.IC

H5

+ *

) +

Ului
X-,
l +

bieo

XT

X i

i *

Xi

+ "*

HJ HS MS HS

Iw

i
ln|r>

MS

*nic;

x.

X.

O.IS

-15 "it

XI

Or

'-4

1 1

l
o>|

"lis
x.
XT
+ffs

x,

Oilln
XT

OlIOB

5|*

I IOT

\r-

x.
i

OD|UI

XT
I r

7T,

#>

,+
M!? 0*15

(.)-

X",
-}(?J(G3

Oi|pJ

PT-,

XT

oolw

b'\U>

inlXT

4-

5|a

S|E2 2|5
X
XT

r|oo 2|n
* 1 XT

5|<

) * "

uH
X,

o|r\3

t-[(J

x-J X

* '*

4-^

0.(|u

Gl-

0|-i

0||-

U,|-

n|r Ullu
XT^

U|U!

X,
1 *

t|0>

x,
i *

x,
i

XT

>

U>1-

M-

vnlwr

ln|c HTS
X,
XT

HS ^IJJ;

b>!C

ofr Ho

x , j XT
i +
w|ln &,|a>

,*

x-,

4^

1-

f
m|oo

rolX,

a i l -

X',

+ '

Hs

o|5

4-

X .vn
1

Wife.

1 -

o
+
10>

XT

A]
1

5
0

<n|2

X,

J-

4-

wl-

b i l -

(n|-

tnl-

Vn|M

(JJlW

o>lw

0|Ul

C9|tP

X,

1
osiro

4-

|-

&fc
XT
**

00|W

x,

HS HS
X",
4 - *

4-

(n
(n|(

1
CSlN

Ho

x,

" 0

( n | -

in

?n

XI

oHn

5l< si^ 5l* 5 1 - 51XT

0.J-

UI|U

HS IS HS

5|oo
7T,

+ -4

XT

09100

5152 5|5
X

x.

H-

U|l

H5

IS "13 |S H!
1

x;
+^

Ui

o|5
X,
I*1

UJ|(M

518 ^IS 5IS

x,

i "
SIS

5IC

X.

0>|-

X",

4/

00|Oi

G>IU

4-

03|-

6>

o.|-

X
1*

1
0B|

5 | - 5 H oili* 5 ^ w h 5|CP 5|ta 515?

X , X * , X ^ x^ x ^ XT Xj^ 7TL
4i
i
!
O
5fc 5|ui 5|o 5+
| ^ 51- 515H

Ha
X
1*

Qt|F

C8|0

&>I4>

,-|n
o|^
X

x;

^
^

1
5*

Cn

4-J

51-

^ 1

5|r*

1 .
Nln

xr

UK
p\Kr\

=IS

x, 1

=B =IS =B

SIS

s!e

-IS u|tt

+
in

ff*

x. 1
4-

m
|

'as
X i

si
7T,

!sla
Inl

x,
+5

1 tnlw
X.
5

00

-vl

ei- =1-

2|i
xl

X,

f09

w(5

S(n
XT

>

XT

ClxT

4-

K|S |S

M|OO

XT

X,

X,

x;

00

2|3 Efo> ^Ifl

4WjOl

XT
r
S|r>

2|oo

x;

X i

*lo

Nj

sl5

^ 'ai*- ai 61 IS|N
TT;
4.0

6>

"A",
1
oil

1 SI-

T,

X*,

+5

1 4-

ln|o

Cn

UIOO

P>

XT

^
i

1 o,|K

Ol|4>

ln|u>

UllUT
inlin

ro

X,

: 1

X,

SI"
XT
1

InlM

i n i -

!0l|^

1 I

S|f >
Ai

X
1

i
in

E|S
XT

H(- E l -

>

X,

CA>|-O

X
i

a-

SI2
1

+*

Sl-

SI1
->1

4-

inl
I
OJ

0^

2IS KIR
XT
1^

a-

KH

->I4
XT

SI5

4 -

n|ra

CO

7.

x,

' ^ 1

SSI5 SIl? lto|0

fall*
X,
x, x ,

+ i

a a* aii
4^

XT
TO

o
+

sis
1

Sh t\~ SI- Sl" $fo W00

X
X,
XJ
X.
+ f
+
41
!
Slfe SIS |SI2 o,_ J| Sl SI* Sl^

SlSIXT

'SI-

rajro

K|- D!|to ]u K|w

ID

ftg

H- !H?
CO

XT

i
in/

2K*

H*
-13

| I
inim
I rnltn
X.

+ p

ID

si-

SiXs
4.0

1
0

CO

At
0

P o w e r s of K.

166

^osjaMoj

od

I 1

*: 2 ^

j BC

CO

1W)
1

su

9rJ H-
u > <tf

sr
+

+
1 r

+
!

-1-

nll^

fl!5

t-i

oj

Sr

SI01
+

-lw>

niNI^

iol+

4-

"*i-

iH
+

ii

L*f

lOJu,

-P
4-

io|io

^ y

OOJU)

3h ?|S
4

ej|5
1

Sl=

sol
4-

li:

iflU
ii)lu

^1rO|U

*|0

41
ol<o

slS
4
"L*:

cvl|o9

4-

^S
KT,|2

-IS
1

^ix

-IS

-IS

S|KJ

<wj

4
ITj

21=

^1^2

o}2

r^
OJJ2

-JC
- | 2

512
+

*:Y

f2|<2
4
0

~|2

<o|!I?
4
^...K
-12

JIS

"115
4

-IS

^(2
4

Y*

sh eh 5l

+
it

t* **
SISI1 talo
1
1

Ml

ol
4

101
*> h o
All
4-

"he
*I
8l- Jth 5H
OJC

*^i

55P

4"

l
uo
CJI
4

ca|

4
^ r
f2|o
1

O
Ml

4-

813:

Inl

1 ^ i -4*
1

^<a|<
ev

^J
U}J

+
2 N

"I*
cO|ui

"I*

inim

a-

2P
-f-

^*niu>

^!h
!
t-

!2N

10
00
+

^ | _&:
I

1 ""

"J
"i^
RSlm

-no

ol

"^
-lo

ui|u)

4
J>L

51- SI"
4

0Q|<*)

^l5

o|2
+

1
0
1^
0)

CO
I

cyj2

Ivr

o|2

0|fflJ

<u|oa

~l

*iiC
Lx
<*
MiSO m|oo
8
1

i^

tO|lfl

w|ui

-110

- | I O

ax
-|UJ

"b<;

2[o

ooln

")\*n

4-

JlQ

COJe^

-te
4

J*)l4

215
1

-12 -12
4-

^-^
1012

"i^

-ICO

+
^Af
oo!2

OJJPJ

**
lOjffl

O
+

eO

CO

HI

A!

CO
+

^co _^
ui
0

T
O

0
4

~l0

4
<U|9

(01")

0|<0
I

-|cu

HN
1

)|<rJ

ta|oi

- I d

_!<vl

-yf
-|M

"IX
-ica

*t
4
1

w>
+
_si
n

- W

Tjd

-|J\!

ca
4

4
jtf

4
O

O
0
O
0
0
O

0
to

<a

JVj

v9

..s

KS

^'

CJ

"~ _

t~-

=
1

^M

<M
4

O
00

eft

tn
4

0
1 \0

4Q

-l0

1*

L....L_ i

(Vj
1

O
1

a
4

-|CM

O
4

"tx

O
4

"^
MlaJ

4
*_Y
oH

4
1 x

1
lid

2l2

O
0 +

-100
4

0OIC0

IJC:
U)|M

l
-

<a
4
Jt

*
tfJllfl

-in
1

0
4-

15
4

Sis

5|2

- 1 *
1

^2| **
5I
-HO
i
"1^

M|Ut

rt|iO

ool0

21"1

5I"1)

HlcJ

n|w

\ ^

CO

CO

CD
IO

<n

CJ

<*

GO

vS

1
l

S|2

" r
OOjcvJ
jca

0
1

U|I

0
4
-iC
00

00

10

0IU>

<)

-xt-

U)

'

5K
IO|U)

>o
1
1^
ih
Kb
I

tflM
1

5S^

O|PJ

?H 511

GO

en

c3

so
1
JC

_kr

eO
1

5
1

51

5
1
o
o

4
-tf
u>

O
10

a
0

0|
-i!
-X.

W)|i^
4

as

1x ^
1

il2
4

U>|J

*JC

4
"IX

1M

rfl

,
l

Sh
2h 5M
1
cJlO

-15

-12 -12
+
^_ir

ol
4

JC

-|in
4

-ISI

4
4i^
"hd
^i*
lOl^t
olo ol* 1

4
3C

^: ^i
0
00

Lx
5l5

Ifl

0
*ix

uyH
+
rt

?i

4-

LJ

-liO

-lo

4
" v

4^v
-Irt

-is

l"0
4

-JXL

-15 -Is
4
4

>I5
1

"IX
*X
"X
"?v
cu|ui OllO u)|in
~JU1

^-1^)
1

4|a>
4

v|0

4
0

r.tl^)

nlo

2ji

<n

">IS - I1S s|$ 5IS

-|U)

1^
wj|ta

JO

o|5

3l->

lOleO

'lx

4
~x

0)
irj

m|s
4

*!*
^P I | 2
-12

r00
5>
<a

4-

cQj>

1^

wl->

51*

4
^ v
J0

*)iS

Lx

?y

a- 1*a- 3H SI- SH 2H
r

ial 2 h

4
4
4
t x
^K
Ik:
H)|lO CMIGO - l a

i h 2b Sh Sl sh 2H
JC

41 00
1 4

' M P S J 9M0J

lad

Sl*>
SI"
1

3| SI- 8l Sl

4<Py

.5

0 +
-_3d

00

r-

IS ^!S "IS -IS -IS -IS

4
" v

-i-

VP

"13 ^jU! -B 2j5 slS 11^

-15

i/D

2t? -R *ii3 "IS "K

+

4wJlH

*!

tO

-IS -IS

<n\oo|c^

?N as

S|2

^ ">:
oOjC
512 rnic[)

u_
O

ATiOM 0

1+
31

~_X

812 !

*L

-Is

4-

en o

U-l
v'

<

2 H

51s
i

^ ^
%\%2x
-IS a IS iS "> IS

lfl|(V
1

ot
Li_l
o
0-.
u_

ol!o
1

*
0

^
1

-f
s-

^
tjJU

<"v

>

*J

0 +

51*

*- rO

10
j

K.

-+-

-la 1 >

G)

|7

3H

to

April

POWERS O F THE GOLDEN SECTION

Si
0

0
0

t-

eO

<n

1966

POWERS OF THE GOLDEN SECTION

167

FIBONACCI AND EUCLID

Walter W. Horner
\ Pittsburgh, Pa.

F o r m a n y c e n t u r i e s n u m b e r addicts have had m u c h fun with the

n u m b e r s g e n e r a t e d by the P y t h a g o r e a n identity
(2mn)

+ (m -n ) = (m +n )

Candida's identity viz.

[mZ+nZ+(m+n)2]

2
= 2

[m4+n4+(m+n)4]

m a y have its m o m e n t s too, e s p e c i a l l y for Fibonacci fans. It is e a s i l y

verified by expanding both s i d e s and obviously holds for all i n t e g r a l
values of m and n.

T h e r e f o r e it holds when m and n a r e con-

secutive Fibonacci or Lucas n u m b e r s . It has its application to g e o m e t r y a s i l l u s t r a t e d by the following p r o b l e m .

"Divide a given s t r a i g h t line into t h r e e s e g m e n t s so that the
s q u a r e on the whole line is twice the sum of the s q u a r e s on the
three segments. "
Solution:

C o n s t r u c t a s q u a r e on line AB.

Let AB (and the r e m a i n -

ing sides) be divided into t h r e e s e g m e n t s which a r e m e a s u r e d by the

second p o w e r s of any t h r e e consecutive Fibonacci n u m b e r s , e. g. ,
2 2 , 3 2 , 5 2 or 4, 9, 25.
A

4
|l6

B
4

25

points of division divide the s q u a r e

81

Then the c r o s s lines joining, the

'

into nine r e c t a n g l e s t h r e e a r e
s q u a r e s and six a r e n o n - s q u a r e s .
Thus we see that the g r e a t s q u a r e

625

25

is twice the sum of the t h r e e s m a l l e r s q u a r e s , i. e c ,

3 8 2 = 2 [16 + 81 + 6 2 5 ] = 1444.

F u r t h e r m o r e the six n o n - s q u a r e s a r e equal in p a i r s and t h e r e f o r e

r e p r e s e n t the six faces of a cuboid.
168

The a r e a and diagonal of the

1966

FIBONACCI AND EUCLID

169

l a r g e s q u a r e a r e equal r e s p e c t i v e l y to twice the a r e a and diagonal of

the cuboid.

This identity can be applied to the a r e a s of c i r c l e s and

s p h e r e s on the line and its s e g m e n t s or to any s i m i l a r polygons d r a w n

on the s a m e .
P r o b l e m i s t s m a y find m a n y applications to g e o m e t r y , e. g. ,
" F r o m the c o r n e r s of an e q u i l a t e r a l t r i a n g l e cut off s i m i l a r t r i angles so that half the a r e a r e m a i n s . "
XXXXXXXXXXXXXXX

(Continued from page 1640 )

,2
b
8

Now if

! - M b =-- 1 3 )

equation (6) h a s the soluti*ons

1 cos s
1 cosh s

L(s) =
Therefore

( cos s 1
(7)

F(s) =

$.z

I cosh s ;
Equation (5) now gives

(8)

u, = 2 cosh A2
k

( cos s

( cosh s )

.I-

1 3

for the hyperbolic cosine a l t e r n a t i v e in equation (6) e

n

Now if A is

chosen to be A r c c o s h -- w h e r e n

is an odd i n t e g e r , it is e a s i l y

shown that the f i r s t t e r m of the right n u m b e r of equation (8) is always

an odd i n t e g e r and hence that u,
of the positive sign.
negative sign.
tively p r i m e .

is not divisible by 2 with the choice

A s i m i l a r r e s u l t holds for

n~ even with the

T h e r e f o r e by the t h e o r e m , the sequence (b) is r e l a The cosine a l t e r n a t i v e in equation (7) leads to a bounded

sequence of integers and therefore is not very interesting.

REFERENCE
1.

Hardy, G H. and Wright, E. M, , '-'An Introduction to the T h e o r y

of N u m b e r s " , Oxford, I960.
XXXXXXXXXXXXXXX

SUMS OF n-th POWERS OF ROOTS OF A GIVEN QUADRATIC EQUATION

N.A. Draim, Ventura, C a l i f . , and Marjorie B i c k n e l l
Wilcox High School , Santa Clara, C a l i f .

The q u a d r a t i c equation whose r o o t s a r e the sum or difference

of the n - t h p o w e r s of the r o o t s of a given q u a d r a t i c equation is of
i n t e r e s t to the a l g o r i t h m a t i s t both for the n u m b e r p a t t e r n s to be seen
and for the implied connection between q u a d r a t i c equations and the
factorability and p r i m e n e s s of n u m b e r s . In this p a p e r , an e l e m e n t a r y
a p p r o a c h which points out r e l a t i o n s h i p s which might o t h e r w i s e be
m i s s e d is used to d e r i v e g e n e r a l e x p r e s s i o n s for the subject equation
and to show how Lucas and Fibonacci n u m b e r s a r i s e as s p e c i a l c a s e s .
Sums of like p o w e r s of r o o t s of a given equation have been
studied in d e t a i l .

F o r a few e x a m p l e s , in this Q u a r t e r l y S. L. Basin

[l~j h a s used a development of W a r i n g ' s formula [ 2 ] for s u m s of like

p o w e r s of roots to w r i t e a generating function for Lucas n u m b e r s and
to w r i t e the kth Lucas n u m b e r as a sum of binomial coefficients.
R. G. B u s c h m a n [ 3 J has used a l i n e a r combination of like powers of
r o o t s of a q u a d r a t i c to study the difference equation u
a r r i v i n g at a r e s u l t which e x p r e s s e s

, = au

+ bu

in t e r m s of a and b.

Paul

F . Byrd [4"j, [ 5 ] h a s studied the coefficients in expansions of a n alytic functions in polynomials a s s o c i a t e d with Fibonacci n u m b e r s .
A n u m b e r of the r e s u l t s in the foregoing r e f e r e n c e s a p p e a r as s p e c i a l
c a s e s of those of the p r e s e n t p a p e r .
A.

THE SUMS OF THE n - t h POWERS

Given the p r i m i t i v e equation f(x) = x

r,

- px -q = 0, with r o o t s

and r ? , the f i r s t p r o b l e m c o n s i d e r e d h e r e is to w r i t e the equa-

tion
i \n

f(x)
with r o o t s

r]

and

=x

, n

n*

*n

- (r1 + r2)x - (rxr2)

= 0

r?.

Solving /the p r i m i t i v e equation for

formula,
170

r,

and r

by the q u a d r a t i c

1966

SUMS O F n - t h POWERS OF ROOTS OF A GIVEN

r x = (P + J?Z + 4q)/2 ,
Since

and r

r 2 = (p - JpZ + 4q)/2

satisfy the equation (x - r )(x - r ) = 0,

2
x

2
- (r

identically, so r , + r
f(x) , having r o o t s

+ r )x + r r

=[(p -

. r,

V/P

=x

= p and r r~ = -q.

+ 4q)/2]

2
2
2
Adding, r , + r ? = p + 2q.
f(x)

- px - q = 0
Next, w r i t e the equation

and r Squaring r o o t s ,

2
r J = [(p + \/p2 + 4q)/2]
r\

171

= (2p 2 + 4q + 2p / p 2 + 4q)/4
= (2p 2 + 4q - 2p / p 2 + 4 q ) / 4 .

The d e s i r e d equation is then

=x

- (p 2 + 2q)x + q 2 = 0 .

3
the coefficient of x for the equation f(x) = 0, by m u l t i 3
3
3
plying and adding, is r + r = p
+ 3pq. Continuing in this m a n n e r ,
the coefficients for x for the equations whose r o o t s a r e higher p o w e r s
of r , and r
can be tabulated as follows:
Similarly,

The sequence for

r, + r

ends when, for a given n, the last

t e r m in the sequence b e c o m e s equal to z e r o .

a s s u m e s the form

2q '

The penultimate t e r m

when n is e v e n a n d npq

'

when n is

odd.
Writing the n u m e r i c a l coefficients from Table 1 in the form of an
a r r a y of m columns and n rows and letting

C(m, n) denote the c o -

efficient in the m t h column and nth row leads to Table 2.

It is ob-

vious that .the i t e r a t i v e o p e r a t i o n in Table 2 to g e n e r a t e additional num e r i c a l coefficients is

C(m, n) = C ( m - 1 , n-2) + C(m, n-1)

which s u g g e s t s a l i n e a r combination of binomial coefficients.

By d i r e c t

expansion,

n ( n - m ) ( n - m - l ) . . . (n-2m+3) _
(m-l)I
"

,n+l-m,
,n-m
m-1 } " (m-l

172

QUADRATIC EQUATION
TABLE 1:

+ r

2 , 2
1
2

April

r* + r

P
P

+ 2q

3 , 3 = 3
T r
P + 3pq
1
2
4
4 , 4
+ 4p q + 2q
r
l + r2 = P

5 , 5

-t- r

+ 5p q + 5:pq'

1
2
6
4
2 2
3
6
6
r
l + r 2 = P + 6p q + 9p q + 2q
r

7 , 7
l + r 2

8,8
l + r2

n .
l

n
2

= p

5
3 2
3
+ 7p q + 14p q + 7pq
+ 8p q + 20p q

n ,

n-Z

+ ... +

+ l6p q

, n(n - 3) n - 4 2 , n(n - 4)(n - 5) n - 6 3

P
q
p
q
~T!
3T
n(n - m)(n - m - 1). . . (n - 2m + 3) n-2m+2 m - 1
( m - 1)!
P
q
TABLE 2:

+ 2q

C(m, n)

SC(m, n)

11

18

14

29

20

16

47

1966

SUMS OF n - t h POWERS OF ROOTS OF A GIVEN

173

F r o m the above t a b l e s ,

where

[a/2]
z^: 1 ) - ( n - i - 1 )
1
1
1
'

n . n
r
l +r2

(A. 1)

[x]

i=0

n-2i i
q

is the g r e a t e s t i n t e g e r l e s s than or equal to x,

and

(n)

is the binomial coefficient

(m-n)l n!
Proof:

By a l g e b r a ,
k
k+1
k+1
, k+1 A k+1
k , k x ,
k t k .
r ,1
+ rn
= (r,
+ r~
+ r n1r29 + Ar l^ 2'
) - ( ^1r 2, + r 1, ^2 )'
= (r^ + r ^ ) ( r i + r 2 ) - r ^ r * "
= p(r

'I

+ r ) + q(r

F r o m Table 1, (A. 1) holds for

+ r

"l)

n = 1, 2, . . . , 8. To p r o v e (A. 1) by m a t h -

e m a t i c a l induction, a s s u m e that (A0 1) holds when n = k and n = k - 1 .

Using the r e s u l t j u s t given and the inductive h y p o t h e s i s ,

[k/2]
k+1 , k+1
r ,1
+ r
=p

2( k :S
. ( k_ii" 1 ) P k-2i q i
i

i=0

[(k-l)/2]
+q
1

k-l-i

k-i-2
P

i=0

Multiplying a s indicated and using the index substitution i-1

k-l-2i i
q
for

the second s e r i e s yields

k+1 , k+1
r
+ r
l
2

[V2]
1

2(k:i) 1 '

(k"i_1)
1

k+l-2i i
q

i=0

0+l)/2]

i
i=0

[2( k :j> - (k;1;11) 1 P

k+l-2i i
q

i in

174

QUADRATIC EQUATION

April

Since the r e c u r s i o n formula for binomial coefficients is

l

m
n'

m
l

n+l;

m+1
n+l'

'

combining the s e r i e s above will yield

[2(k+i-i}

for the coefficient of p

and for

q .

(k-i}]

Considering the s e r i e s for

k even

k odd leads to

[(k+l)/2]
k+1 ,

rx

k+1

+r2

'

2(k+l-ij

_ (k-i} '

k+l-2i i
q

i=0

so that (A. 1) follows by m a t h e m a t i c a l induction.

An e x p r e s s i o n eq-

uivalent to (A. 1) was given by Basin in [ ! ]

Using formula (A. 1), we can w r i t e the d e s i r e d equation
,. n
2
n
n.
. vn n n
f(x) = x - (r x + r 2 ) x + (-1) q = 0 .
2
Example:
Given the equation x - 5x + 6 = 0, w r i t e the equation whose r o o t s a r e the fourth powers of the r o o t s of the given equation, without solving the given equation. In the given equation, p = 5,
q = 6. F r o m Table 1 or formula (A. 1),
r^ + rjj = (5) 4 + 4(5) 2 (-6) + 2 ( - 6 ) 2 = 97 ,
4

2
= 0 . As a check, by factoring x -5x + 6 = 0,
4
4
4
2
4
r ] = 2 and r_ = 3, giving us f(x) = (x - 2 )(x - 3 ) = x - 97x + 6 .
R e t u r n i n g to Table 2, the s u m m a t i o n of the coefficients by rows
yields the s e r i e s 1, 3, 4, 7, 11, 18, 29, 47, . . . , the s u c c e s s i v e
so f(x)

=x

- 97x + 6

Lucas n u m b e r s defined by L, = 1, L n = 3, and L = L , + L n .

7
1
2
n
n-1
n-2
The g e n e r a l equation for r, + r ? r e d u c e s to the n u m e r i c a l values of
2
C(m, n) of Table 2 for p = q = 1 in the p r i m i t i v e equation x - px -q =
c

0, which b e c o m e s x - x - 1 = 0 with r o o t s a = (1 + /fT)/2 and

(3 = (1 - \/~5)/2. This yields an e x p r e s s i o n for the nth Lucas n u m b e r , since

L = a + 6 ;
n

1966

SUMS OF n - t h POWERS OF ROOTS OF A GIVEN

(A. 2)

L^ = 1 + n + n ( N - 3 ) / 2 ! + n ( n - 4 ) ( n - 5 ) / 3 ! + n ( n - 5 ) ( n - 6 ) ( n - 7 ) / 4 !

[n/2]
1

2(n:i) -

175

(n'Yly

i=0
F o r e x a m p l e , L, = 1 + 6 + 6(3)/2 + 6(2)(l)/6 = 18.
2
Also, the equation whose r o o t s a r e the nth roots of x - x - 1 = 0
is x - L x + (-1) = 0. ( E s s e n t i a l l y , equation (A. 2), a s w e l l a s (A. 3),
(B. 2), (B 0 3), a p p e a r s in [ 3 ] by B u s c h m a n . )
An a l t e r n a t e e x p r e s s i o n for L is obtained by e x p r e s s i n g
r

+r

in t e r m s of p and A , w h e r e A = \/p

+ rq e

Now,

n
n
T i _LA\/-?1 n
o-n/n.
~l
. n ( n - l ) n~2 2 . n ( n - l ) ( n - 2 ) n - 3
r x = L(P +A )/2J = 2 (p + np
A +
3/
'p
A + - ^ T " P

n r,
\/oin
r 2 = [(P - A )/2J
which, for
(A. 3)

= 2o~

/ n n n _ lA.
(P - P

p = q = 1, A = v5,
L

+.

n ( n - l ) n-2 2 , n ( nn - l ) ( n - 2 ) n - 3
"^T~- P
A +-^r
P

i m p l i e s , on adding,

= 2 1 ~ n ( l - 5 n ( n - l ) / 2 I + 5 2 n(n~ l ) ( n - 2 ) ( n - 3 ) / 4 ! + . . . )

[n/2]
1

?n-l
Z

B.

_i, n,

5^;.)
x

2r

i=0

THE D I F F E R E N C E S OF THE nth POWERS

Now let us t u r n to the equation whose r o o t s a r e

where r

and r

a r e the r o o t s of the p r i m i t i v e equation x

Ct

r
2

and

-r

-px -q = 0

Since
r

"

( p

p 2

r q ) / / 2

"

( p

" Vpl

4 q ) / / 2

the equation whose r o o t s a r e r , and ~r is x

2
and f(x) = x - A x + q = 0. S i m i l a r l y ,

*\-x\=

/P2

+ 4(

- yp

1)/ 2 ] 2 - [(P - V/P 2

p 2

4 q

+ rq x - r , r

+ 4

<1>/2] 2 =

'

= 0,

3,
+
3,
+

176

April

QUADRATIC EQUATION

Continuing in this m a n n e r ,

and a s s e m b l i n g r e s u l t s , we have

TABLE 3
n
l "

n
2

A.

PA

(p

( p 3 + 2pq)A

(p4 + 3p2q + q2)A

(p5 + 4p3q + 3pq2)A

( p 6 + 5 p 4 q + 6 p 2 q 2 + q 3 )A.

, n-1
. _. n - 3 ^ ( n - 3 ) ( n - 4 ) n - 5 2
(p
+ (n-2)p
q+-i
JJ
' p
q + ...

+q)A

( n - m - l )Li( n - m - 2 )' , . , ( in - 2 m ) L p n - 2 m - l q m .) A
.
+, 1_

mi

The sequence for

r1 - r ?

as given in Table 3 ends, for a given n,

on a r r i v i n g at t e r m z e r o .
Writing the n u m e r i c a l coefficients from Table 3 in the form of
an a r r a y of m columns and n r o w s , and letting C (m, n)
the coefficient in the m t h column and nth

TABLE 4:
m

row,

C (m, n)
A

% C A ( m , n)

13

10

21

15

10

5
8

34

denote

1966

SUMS OF n - t h POWERS OF ROOTS OF A GIVEN

177

It is obvious that the i t e r a t i v e p r o c e s s to g e n e r a t e additional n u m e r i c a l

coefficients r e s t s on the r e l a t i o n
C A ( m ? n ) = C A ( m - l , n-2) + C A ( m , n-1)
the s a m e a s for Table 2

The coefficients g e n e r a t e d a r e a l s o the c o -

efficients a p p e a r i n g in the F i b o n a c c i polynomials defined by

f n + 1 (x) = x f j x ) + f n _ 1 ( x ) ,

f 1 (x) = l ,

f 2 (x)

( F o r further r e f e r e n c e s to Fibonacci p o l y n o m i a l s ,

= x

see Byrd, [ 4 ] ,

p.17.)
Tables 3 and 4 lead to

(B.l)

rj-rj

= A

l>/2]

(""j"1) P ^ ' V

j=0
which is proved s i m i l a r l y to (A. 1)
The equation whose r o o t s a r e
4,-u

a r e the r o o t s of

r\

- px - q = 0,

r ' and - r ? , w h e r e
n

is

f(x) A = x

- (r

r
n

and
J_

- r )x + q
i

r
n

=0.

We use ( B . l ) to solve the following: Given the equation x - lOx +

21 = 0 with unsolved r o o t s r , and r ? , w r i t e the q u a d r a t i c equation
3
3
whose r o o t s a r e r , and - r ? F r o m the given equation, p = 10, q = - 2 1 ,
r

and A = / 1 0

- 4(21) = 4,
r

l " r2

Using ( B . l ) or Table 3,
= (p2 + q)

= (1

2 " 21)4

= 3l6

3
2
3
Hence f (x) A = x - 3 l 6 x + (-21) = 0 . As a check, the r o o t s of the p r o b 3
3
3
2
lem equation a r e r ] = 7, r = 3, so f(xL = (x - 7 )(x + 3 ) = x - 3 l 6 x 3
21 = 0.
The s u m m a t i o n of t e r m s in the s u c c e s s i v e rows of Table 4 is s e e n
to yield the sequence of F i b o n a c c i n u m b e r s , defined by F , = F ? = 1,
F

= F , + F o8 This r e s u l t can be d e m o n s t r a t e d by7 substituting

&
n
n-1
n-2
p = q = 1 in the g e n e r a l e x p r e s s i o n (B. 1) and dividing by A = / 5 , for
the left hand m e m b e r b e c o m e s the Binet form, F = (a11 - p")/ ST. In
general,

178

QUADRATIC EQUATION

(B.2)

April

= 1 + (n-2) + (n-3)(n-4)/2! + (n-4)(n-5)(n-6)/3! + . . .

j=o
the sum of t e r m s in r i s i n g diagonals of P a s c a l ' s t r i a n g l e .

Also, the

q u a d r a t i c equation whose r o o t s a r e a and -(3 , nth p o w e r s of the

2
r o o t s of the q u a d r a t i c equation x - x - 1 = 0 with r o o t s
a = (1 + / 5 ) / 2

is x 2 - / 5 " F x - 1 = 0.

and (3 = (1 - / 3 ) / 2 ,

Expressing

r, - r ?

t e r m s a s we did for

in t e r m s of only p and A and a s s e m b l i n g

r, + r

in (A. 3), we a r r i v e at an a l t e r n a t e ex-

p r e s s i o n for the nth F i b o n a c c i n u m b e r ,

(B. 3) F

= 2 1 ~ n (n -I- 5 n ( n - l ) ( n - 2 ) / 3 ! + 5 2 n(n- l)(n-2)(n~3)(n-4)/5! +. . .

[(n-l)/2]
X

2 "

R 11

5 <(

1
i=0

2i+lJ

REFERENCES
le

S. L0 Basin,

"A Note on W a r i n g ' s F o r m u l a for Sums of Like

P o w e r s of R o o t s " ,

F i b o n a c c i Q u a r t e r l y 2:2, April, 1964, pp.

119-122.
2.

L. E. Dickson, F i r s t C o u r s e in the T h e o r y of Equations, John

Wiley, 1922.

3.

R. G. Buschman,
ials,

" F i b o n a c c i N u m b e r s , Chebyshev P o l y n o m -

Generalizations

and Difference E q u a t i o n s " ,

Fibonacci

Q u a r t e r l y 1:4, D e c , 1963, pp. 1-8.

4.

P a u l F . Byrd,

"Expansion of Analytic Functions in P o l y n o m -

i a l s A s s o c i a t e d with Fibonacci N u m b e r s " , F i b o n a c c i Q u a r t e r l y ,

1:1, F e b . 1963, pp. 16-29.
5.

P a u l F . Byrd,

"Expansion of Analytic Functions in T e r m s In-

volving Lucas N u m b e r s or S i m i l a r Number S e q u e n c e s " , F i b onacci Q u a r t e r l y 3:2, A p r i l , 1965, pp. 101-114.

XXXXXXXXXXXXXXX

DETERMINANTS INVOLVING Kth POWERS FROM SECOND ORDER SEQUENCES

David A. Klarner
University of Alberta, Edmonton

INTRODUCTION
Let a
be a sequence of complex n u m b e r s satisfying the diff e r e n c e equation
a ,- = a a . . - (3 a
n+2
n+1
n

d!

for

w h e r e a and (3 a r e fixed complex n u m b e r s ,

we define
k

k
a ._
n+1

a
n
(2)

Ak(an)

n+l

n+k

n = 0, 1, . . .

for such a sequence

n+k
a

n+2

k
n+k+l

n+k+l

for

n = 0, 1, .

n+2k

It is the p u r p o s e of this note to prove

(3)

W'^'^'^VV

and give e x a m p l e s of the r e s u l t

A D I F F E R E N C E EQUATION FOR f a k ] :
*- n J
Z
Let ( 1 - 91x)(l- 9 x) = 1-ax + (3x , so that 0 + # = a and
# 9 = p, and a s s u m e

9 /

0 .

C a r l i t z [ 3 j has proved that

oo

(4)

Pk(x) / qk(x) = 2
n=0

where
179

n X"

180

DETERMINANTS INVOLVING Kth

qk(x) = n

(5)

April

POWERS

(i- e\ ez'x x)

i=0

and p (x) is a polynomial of d e g r e e l e s s than the d e g r e e of q (x).

Letting

(6)

k(x)=

- Yl

k+l-i(k)xl

i=l
k

(the constants

a.( )

a r e polynomials s y m m e t r i c in

and

t e r m i n e d by (5)) we see after multiplying through (4) with q (x)

given in (6)) and equating coefficients of x

de(as

in the right and left m e m -

b e r s that
(7)
for

k
j.1 xi " a ik+1
xi ( k )
n+k+1

n = 0,1,...

k + a k
xi
i k( )
n+k

k
k
xi - , + . . . + CL (k)a
n+k-1
I n

We a l s o know from (*>) and (6) that

/ i \ k + 1 / ( *n a$ )x l+2+. . . +k
- a j t /ik )v = (-1)
x
2

(8)

al(k) = ( - D k + 2

(k+1>/2

Now let k be a fixed n a t u r a l n u m b e r and c o n s i d e r for

(9)

or

n > 0,

( - l ) k ( - l ) k + 2 ( 3 k ( k + 1 ) / 2 Ak(an) = ( - l ) k a^k) Ak(an)

n+1

n+2

k
*n+2

k
Si+3

n+k+1

a n+k+2
"".. . 0

/i \
k

n+k

i< >

n+k+l

l(k)an+l

n+2k .

l(k)an+k

VW

The second equality in (9) follows since we have i n t e r c h a n g e d

u m n s and multiplied the l a s t column by a, (k).

k col-

The l a s t equality fol-

lows since a p p r o p r i a t e m u l t i p l e s of the f i r s t columns can be added to

1966

F R O M SECOND ORDER SEQUENCES

181

the l a s t column to m a k e it the last column of A, (a i n ) ; the " a p p r o k n+1

^
p r i a t e m u l t i p l e s " a r e the a.(k) given in (7).
Thus, we have shown
^
^
( . 1 ) k (_1)k+2 pk(k+l,/2 ^
= pk(k+l)/2 ^
= V a n + i )
so that (3) can be proved by induction on n.
As a c o r o l l a r y to (3) we note that if ) a v s a t i s f i e s (1), then
ja

I , w h e r e q and p a r e non-negative i n t e g e r s , is a second

o r d e r sequence a s well; in fact,

(10)

q(n+2) + p " ( * 1

+6 )

P 4 a qn+p

q(n+l) + p

Hence we can r e w r i t e (3) to obtain

(11)

, = p1nk(k+1)/2A,(a )

AJa

qn+p)

E X A M P L E S INVOLVING THE FIBONACCI SEQUENCES

When

is the Fibonacci sequence -J F

0,1,1,2,

(3 = -1 in (3) so that we have

(13)

(14)

n+1

(12)

3
! Fn
3
F n+1
3
F.n+2
3
F
n+3

n+1

'n+2

.2
n
2
n+1
2
n+2

2
n+1
2
n+2
2
n+3

n+1
3

n+2
F

n+2

n+3

n+3
3
F
n+4

F
^n+4
3
F
n+5

(-1) 1

n+2
2
n+3
2
n+4f

= (-1)

1)

i-

n+1

3n

M-Dn+1 2

n+3
3
F'
n+4
F3
^n+5
3
F
n+6

(-1)

6n

0 1

1 1

27

1 8

27

125

27 125 512

36

182.

DETERMINANTS INVOLVING Kth POWERS

April

The r e s u l t in (12) is well known, B r o t h e r Alfred p r o p o s e d (13)

a s a p r o b l e m in the v e r y f i r s t i s s u e of the Fibonacci Q u a r t e r l y [l^\f
and E r b a c k e r ,

Fuchs and P a r k e r p r o p o s e d (14) in a l a t e r i s s u e |~5"J.

If we redefine

aA = F 1 } a. = F _ , . . . we have ! a I = i u | in
0
1 1
2
^ n$
^ nj
the s t a n d a r d notation; fixing q = 2 and p = 1 in (11) we obtain for
k = 1 and 2,
(15)

A (u

(16)

l 2n+1)

= A

1 ( V = -1

2 ( u 2 n + 1 ) = A 2 ( U 1>

18

r e s p e c t i v e l y ; on the other hand if we fix q = 2 and p = 0 in (11) we

have for

k = 1 and 2,

(17)

A (u

(18)

A (u

2n)

= A

l(u0)

2 2n>

= A

2<V =

respectively.

= 1

Together (l6)and (18) imply

u

(19)

n+2

n+4

2
n+2

2
n+4

2
n+6

2
n+4

2
n+6

2
n+8

= (-l)n+1

18

which has a l s o been p r o p o s e d a s a p r o b l e m by B r o t h e r Alfred [ 2 ] .

AN E X A M P L E INVOLVING A SEQUENCE OF POLYNOMIALS
L o r c h and M o s e r [8~j p r o p o s e d that one prove
sT

n+1

(20)

v
n+1
where

(21)

for n = 0 , 1 , 2 , . . .

n+2

v n = 1 and

E
v=0

.n+v. v
(
) x

for n = 1,2,

FROM SECOND ORDER SEQUENCES

183

In proving (2), C a r l i t z [ 4 ] proved

V n + 2 = (x+2) v n + 1 - v n

for n = 0 , 1 , 2 , . . .

, we c a n p r o v e (20) and obtain g e n e r a l i z a t i o n s by using (3).

and 2 we have r e s p e c t i v e l y ,
A (v

l J

For

=A

1<V = X '

A 2 ( v n ) = A 2 (v Q ) = 2x 3 (x+2) 2

A second g e n e r a l i z a t i o n of this p r o b l e m was a l s o given by Gould

REFERENCES
B r o t h e r U. Alfred, P r o b l e m H - 8 , Fibonacci Q u a r t e r l y , F e b . ,
1963, p. 48.
B r o t h e r U. Alfred, P r o b l e m H-52, F i b o n a c c i Q u a r t e r l y , F e b . ,
1965, p . 44.
L. C a r l i t z , "Generating Functions F o r P o w e r s of C e r t a i n Sequences of N u m b e r s " ,

Dtike M a t h e m a t i c a l J o u r n a l ,

vol. 29,

(1962) No. 4, pp. 521-538.

L. C a r l i t z , Solution to P - 3 3 , Canadian M a t h e m a t i c a l Bulletin,
vol. 4, (1961) No. 3, pp. 3 1 0 - 1 1 .
J. E r b a c k e r , J. A. F u c h s , F . D. P a r k e r , P r o b l e m H - 2 5 , F i b onacci Q u a r t e r l y , D e c , 1963, p. 47.
H. W. Gould, "A G e n e r a l i z a t i o n of A P r o b l e m of L. Lorch and
L. M o s e r " ,

Canadian M a t h e m a t i c a l Bulletin,

vol. 4,

(1961)

No. 3, pp. 3 0 3 - 5 .
D. A. K l a r n e r , C h a p t e r s 3 and 4, "On L i n e a r Difference Equations" (Master's Thesis),

University of A l b e r t a ,

Edmonton,

A l b e r t a , Canada, pp. 3 0 - 4 9 .
L. L o r c h and L. M o s e r ,

P r o b l e m 33, Canadian M a t h e m a t i c a l

Bulletin, vol. 4, (1961), No. 1,

p. 310.

XXXXXXXXXXXXXXX

A GENERALIZED LANGFORD PROBLEM

Frank S. G i l l e s p i e and W.R. Utz
\ University of M issour i, Columbia, Mo.

Let n > 1 be an i n t e g e r and c o n s i d e r the i n t e g e r s 1, 2, 3, . . . ,

n. The sequence a , , a OJ a 0 , . . . , a~> is said to be ;a p e r f e c t s^e^
1
2
3
2n
quence for n if each of the i n t e g e r s 1, 2, 3, . . . , n o c c u r s in the
sequence exactly twice and the i n t e g e r

i is s e p a r a t e d in the sequence

by exactly i e n t r i e s .

1 7 1 2 6 4 2 5 3 7 4 6 3 5

p e r f e c t sequence for

For example,

7. C. D. Langford [ 2 ] posed the p r o b l e m of d e -

t e r m i n i n g all n having a p e r f e c t s e q u e n c e .
Priday

is a

3 and Roy O. Da vies

only if, n is of the form

It was shown by C. J.

1 that n has a perfect sequence if, and

4 m - 1 or 4 m .

For

n = 3, 3 1 2 1 3 2 i s

the only perfect sequence except for the s a m e sequence in r e v e r s e o r d e r

and for

n=4, 4 1 3 1 2 4 3 2

is the only p e r f e c t sequence except for

the s a m e sequence in r e v e r s e o r d e r .
25 p e r f e c t s e q u e n c e s for 7,

According to Davies t h e r e a r e

He stated the p r o b l e m ,

as yet unsolved,

of finding a function giving the n u m b e r of perfect s e q u e n c e s for

n of

the form 4m - 1 or 4 m .
In this note we define a g e n e r a l i z e d p e r f e c t s - s e q u e n c e for the
integer

tegers

1, 2, 3, . . . , n o c c u r s exactly

> 1 to be a sequence of length sn in which each of the in-

o c c u r e n c e s of the i n t e g e r
sequence for

i there are

s t i m e s and between any two

i entries.

Thus, a p e r f e c t

n is a g e n e r a l i z e d p e r f e c t 2 - s e q u e n c e c

The a u t h o r s a r e unable to d i s c o v e r an n for which t h e r e is a

g e n e r a l i z e d p e r f e c t s - s e q u e n c e for

s > 2 and pose a s a p r o b l e m the

d e t e r m i n a t i o n of all s and n for which t h e r e a r e g e n e r a l i z e d p e r f e c t

s - s e q u e n c e s for

n.

The following p a r t i a l r e s u l t is given in c a s e

of proof b e c o m e s tedious for l a r g e

s = 3.

The method

n but could be settled for any given

n on a m a c h i n e .
Theorem.

T h e r e is no g e n e r a l i z e d p e r f e c t 3 - s e q u e n c e for

n = 2, 3, 4, 5, 6.
Proof.

The c a s e n = 2 is t r i v i a l ,
184

1!966

A GENERALIZED LANGFORD PROBLEM

Consider

the c a s e

p e r f e c t 3 - s e q u e n c e for
we m u s t h a v e 3, a , ,

n = 3.

n = 3.

IBS-

A s s u m e that t h e r e is a generalized

Beginning with the f i r s t 3 in the s e q u e n c e

a , a , 3.

T h e r e a r e 9 e l e m e n t s in the s e q u e n c e

i n c l u d i n g a n o t h e r 3 h e n c e t h e e n t i r e s e q u e n c e m u s t b e of t h e f o r m
3, a ^ a^, a^, 3, t ^ , b ^ , b ^ , 3
T h e f i r s t o c c u r e n c e of 2 i n t h e s e q u e n c e i s a t
a~ = b , = 2 o r

.
B.

or

a^

hence either

a~ = b~ = 2 b u t i n n e i t h e r c a s e i s t h e r e r o o m f o r a n o t h e r

2 and so t h e r e is no g e n e r a l i z e d p e r f e c t 3 - s e q u e n c e for
Now,

let

n = 4.

n = 3.

If t h e d e s i r e d s e q u e n c e i s p o s s i b l e ,

beginning

w i t h t h e f i r s t 4 i n t h e s e q u e n c e w e h a v e 4, a, , a~, a , a . , 4, b n , b ~ , b O J b ., 4 .
^
1 2
3 4
1 2 3 4
B e c a u s e of t h e p o s i t i o n s

of t h e 4 f s , a , , a~, a . , b , , b

a r e not 3 h e n c e

a~ = b ? = 3 a n d t h e s e q u e n c e e i t h e r b e g i n s o r e n d s w i t h a 3 .

Consider

the c a s e
4, a 1 ? a 2 > 3, a 4 ? 4, b ^ 3, b ^ , b^, 4, 3
the a l t e r n a t e case is s i m i l a r .

B e c a u s e of t h e s p a c e s a l r e a d y o c c u p i e d

b y t h e 3 ' s a n d ,4's i t i s n o t p o s s i b l e to p u t t h e 2 ' s in t h e s e q u e n c e a n d

so

n = 4 is i m p o s s i b l e .
In c a s e

n = 5, w e m u s t h a v e t h e s u b s e q u e n c e
5, a , , a~, a^,, 3-A> a c j ^> b, , b~, b~, b ., b,., D

in the p r o p o s e d s e q u e n c e .
If

It i s o b v i o u s t h a t

a , , a ? , a-

cannot be a 4.

a~ = 4, t h e s e q u e n c e i s

(1)

4, c x , 5 , a l f a 2 , 4, a 4 , a 5 , 5, b ^ 4, b 3 , b 4 , b 5 , 5

or has
(2)

5, a 1 ? a 2 , 4, a^, a g , 5, b ^ 4 , b 3 , b ^ b g , 5, 4

as a subsequence.
(3)

If a 4 = 4, there is a subsequence

4, 5, a x , a z , a ^ 4, a^, 5, by b^9 4, b 4 , bg, 5

or the e n t i r e sequence is
(4)

5 , a x , a 2 , a^, 4, a 5 > 5 , b ^ b ^ , 4 , b 4 > b g , 5 , d ^ 4

186

A GENERALIZED LANGFORD P R O B L E M

April

F o r sequence (1), it is d e a r that one m u s t have a, = a- = b

= 3

hence a = b 1 = b . = 2 but this is i m p o s s i b l e since one cannot have

4
1 4
^
c, = a~ = b - = 1. The a r g u m e n t for sequence (2) is the s a m e ,
F o r the sequence (3), the only possible choices for 3 m a k e
a 0 = b, = b c = 30 This done, the only choices for 2 m a k e a~ = a- = b~ = 2
D

but this r e q u i r e s that a, = b . = 1 which is i m p o s s i b l e .

The a r g u m e n t

for the sequence (4) is the s a m e and it is seen that the c a s e n = 5 cannot o c c u r .
The c a s e n = 6 is t r e a t e d s i m i l a r l y .

The d e t a i l s a r e n u m e r o u s

and will be omitted.

The a u t h o r s a r e indebted to the r e f e r e e for the following t h e o r e m .
Theorem.

T h e r e is no g e n e r a l i z e d p e r f e c t s - s e q u e n c e

for

n < So
Proof.
s-1

There are

p a i r s of adjacent

s t e r m s equal to n, and between each of the

n's

is an i n t e r v a l of length n.

The total length,

s + n ( s - l ) , m u s t not be g r e a t e r than sn, which i m p l i e s

n > s.

REFERENCES
1.

Davies, Roy O. :

On Langford's p r o b l e m (II), Math. Gaz. 43

(1959), pp. 253-255.

2.

Langford, C. D. : P r o b l e m , Math. Gaz 0 42 (1958),

3.

P r i d a y , C. J . : On Langford's p r o b l e m (I), Math. Gaz. 43 (1959),

pp. 2 5 0 - 2 5 3 .
xxxxxxxxxxxxxxx

p. 228.

A LOWER BOUND FOR MAXIMUM ZERO-ONE DETERMINANTS

B.L. Foster
Denver Research Center, Marathon Oil Company
Littleton, Colo.

What is the l a r g e s t p o s s i b l e d e t e r m i n a n t of o r d e r n, if z e r o and

one a r e the only e n t r i e s a l l o w e d ? This question, last posed by H a r a r y
("l~j , s e e m s difficult,,
Williamson [2"] obtained the values 1, 1, 2, 3, 5, 9 for n = 1, 2,
3, 4, 5, 6; and proved the g e n e r a l p r o b l e m equivalent to a s i m i l a r
H a d a m a r d question with allowed e n t r i e s 1 and - 1 .
Cohn [ 3 ] d e r i v e d an a s y m p t o t i c lower bound,
( n + 1 -0(n+l) / 2 n

where

c is any positive n u m b e r .

The upper bound

(n+1)(n+l)/2/2n

follows from H a d a m a r d ' s inequality [ 4 l , applied to the 1, -1 v e r s i o n

of the p r o b l e m .
C l e m e n t s and Lindstrorn [5 J have announced the lower bound
(n+l)K/Zn .
where

K = (n+1 )(1-(log 4 / 3 ) / l o g ( n + l ) ) / 2 , and the l o g a r i t h m s a r e b a s e

two.
In this note, I show that the Fibonacci sequence 1, 1, 2, 3, 5,
8, . . . is a lower bound for the sequences of m a x i m u m z e r o - o n e d e t e r m i n a n t s . Also, I c o m p a r e this bound with the C l e m e n t s - L i n d s t r o r n
bound.
Theorem:

The m a x i m u m z e r o - o n e d e t e r m i n a n t of o r d e r

at l e a s t a s l a r g e a s the n

n is

Fibonacci n u m b e r .

This is proved by exhibiting zero-one matrices whose determinants are the Fibonacci numbers.
Let-.a(n) be the row v e c t o r with n e n t r i e s which a r e a l t e r n a t e l y
one and z e r o , s t a r t i n g with one. Con
Consider the n
order matrix
187

188

April.

A LOWER BOUND FOR MAXIMUM

a(n)
1
0

a(n-l)
1
0

F(n) =

a(2)
1

a(l)

F o r example,
/ 1 0 1
1 0

F ( l ) = (1), F(2)

, and

1 10

F(3)

1 1

0 11

Notice that det F ( l ) = 1, det F(2) = 1, and det F(3) = 2 .

Expanding det F(n) by the f i r s t column gives det F(n) =
/0 a(n-2)
1 a(n-2)
det

a(n-3)

det I .

a(2)

a(2)
a(l)l

a(n-2)
1

a(n-3)

= det F ( n - l ) - (-l)det

a(l)

0
a(2)
1

= det F ( n - l ) + det F ( n - 2 ) .
Fibonacci sequence.

Therefore,

a(l),

the sequence det F(n) is the

To c o m p a r e this bound with the C l e m e n t s - L i n d s t r o m bound, e x a m i n e the following t a b l e .

1966

ZERO-ONE DETERMINANTS

| det F(n)

189

13

21

34

1. 1

1.7

2.8

5. 2

10. 1

21.1

46. 3

10

11

12

13

14

55

89

144

233

377

610

4, 669

13, 122

[jn+l) /2

107.2

.8

.9

259.5

654.9

1, 717.7

15 j

If n is g r e a t e r than 8, the C l e m e n t s - L i n d s t r o m bound is b e t t e r .

F o r speciaL n, still b e t t e r bounds can be found.

One of Cohn's

i n e q u a l i t i e s ("3 | b e c o m e s , for z e r o - o n e d e t e r m i n a n t s ,
M(mn-l) > z*- 1 ** 11 - 1 ) [ M ( m - l ) ] n [ M ( n - l ) ] m
where

M(i) is the m a x i m u m d e t e r m i n a n t of o r d e r

i.

If m n - 1 = 14

and m n - 1 = 15, then

M(14) > 6, 912,

and

M(15) > 131,072

The n u m b e r s in the table above w e r e bought from Diane K. Middents for 2 p a l i n d r o m e s .

REFERENCES
1.

F . H a r a r y , R e s e a r c h P r o b l e m 1, Bulletin of the A m e r i c a n Math

e m a t i c a l Society 68 (1962) no. 1, p. 2 1 .

2.

J. Williamson, " D e t e r m i n a n t s whose e l e m e n t s a r e 0 and 1, "The

A m e r i c a n M a t h e m a t i c a l Monthly 53 (1946), pp. 4 2 7 - 3 4 .

3.

J. H. E. Cohn, "On the value of d e t e r m i n a n t s , " P r o c e e d i n g s of

the A m e r i c a n M a t h e m a t i c a l Society, 14 (1963) no. 4, pp. 5 8 1 - 8 8 .

4.

J. H a d a m a r d ,

"Resolution d'une question r e l a t i v e aux d e t e r m i -

n a n t s , " Bulletin des Sciences M a t h e m a t i q u e s (2), Vol. 17, 1893,

pp. 240-246.
5.

G. F . C l e m e n t s and B. L i n d s t r S m ,

"A sequence of ( l ) - d e t e r -

m i n a n t s with l a r g e v a l u e s , " Notices of the A m e r i c a n M a t h e m a t i cal Society 12 (1965) no. 1, pp. 7 5 - 76.
6.

Don W a l t e r s , A m e r i c a n M a t h e m a t i c a l Atfonthly,
Problem E-834.
xxxxxxxxxxxxxxx

J u n e - J u l y 1949

ELEMENTARY PROBLEMS AND SOLUTIONS

Edited by A . P . Hillman
University of New Mexico, Albuquerque, New Mexico

Send a l l c o m m u n i c a t i o n s r e g a r d i n g E l e m e n t a r y P r o b l e m s and
Solutions to P r o f e s s o r A. P . Hillman,

D e p a r t m e n t of M a t h e m a t i c s

and S t a t i s t i c s , U n i v e r s i t y of New Mexico, Albuquerque, New Mexico.

E a c h p r o b l e m or solution should be submitted in legible form,

pref-

e r a b l y typed in double spacing, on a s e p a r a t e sheet or s h e e t s in the

f o r m a t u s e d below. Solutions should be r e c e i v e d within t h r e e months
of the publication d a t e .
B-88

Proposed by John Wessner, Melbourne,

Let

L^, L^, L, L , , . . .
0 L 4 o

Florida

be the Lucas n u m b e r s 2, 3, 7, 18, . . .

Show that
L2k= 2(-l)k(mod5)
B-89

Proposed by Robert S. Seamons, Yakima Valley College, Yakima,

Washington

Let F

and L be the n - t h Fibonacci and n - t h Lucas n u m n

n
ber } r e s p e c t i v e l y . Let [x"j be the g r e a t e s t i n t e g e r function.
Show
that L 0
= 1 + LT vJ / I T F 0 J1 for a l l positive i n t e g e r s m .
2m
2m
B-90

Proposed by Phil Mana, University

of New Mexico, Albuquerque,

Let b , b , . . . be the sequence 3 , 7 , 4 7 , . . .

2
r e l a t i o n b ., = b - 2 . Show that the r o o t s of
n+1
n

New Mexico

with r e c u r r e n c e

x 2 - 2b x + 4 = 0
n
a r e e x p r e s s i b l e in the form
B-91

c + d /"F, w h e r e

Proposed by Douglas hind, University

If F

i s the n - t h

c and d a r e i n t e g e r s .

of Virginia, Charlottesville,

F i b o n a c c i n u m b e r , show that
oo

]r<i/F.)
190

Virginia

1966

ELEMENTARY P R O B L E M S AND SOLUTIONS

191

c o n v e r g e s while
oo

(l/lnF.)
j=3
diverges*
B-92

Proposed by J.L. Brown, Jr., The Pennsylvania State University

Let

(x, y)

denote the g. c, d. of positive i n t e g e r s x

and

y.

Show that (F , F ) = ( F , F . ) = ( F , F , ) for all positive i n t e m

n
m
m+n
n m+n
^
g e r s m and n.
B-93

Proposed by Martin Pettet,

Toronto, Ontario,

Canada

Show that if n is a JTpositive X-p r i m e , L n = 1 (mod n).

Is the con-

verse true ?
SOLUTIONS
The deadlines for submitting solutions to the E l e m e n t a r y P r o b l e m s Section have proved to be u n r e a l i s t i c and a r e being changed as of
this i s s u e .

We take this opportunity to list s o m e e r r a t a and s o m e of

the s o l v e r s whose solutions w e r e r e c e i v e d after copy for this section

went into production,
ERRATA
B-33

The solution p r i n t e d on page 235, Vol. 2, No. 3, was submitted

by C h a r l e s R. Wall,

Texas C h r i s t i a n U n i v e r s i t y , F t . Worth, T e x a s .

This p r o b l e m was a l s o solved by John H. Ha It on, B. Litvack, and the

proposer.
B-39

In the note after the solution to B-39 on page 327, Vol. 2, No.

4, the r e f e r e n c e

to the inequality a

< F

< a

for

n > 1 in An

Introduction to the T h e o r y of N u m b e r s , b y N i v e n a n d Z u c k e r m a n ,
state that their initial conditions on the F
B-57

should

a r e F . = 0 and F 0 = 1.
n
1
2
On the next to l a s t line of the solution on page 160, Vol. 3, No.

2, the second "= " sign should obviously be a " > " sign.

192

ELEMENTARY PROBLEMS AND SOLUTIONS

April

SOLUTIONS RECEIVED A F T E R PRODUCTION DATE FOR

P R O B L E M SECTION
Problem

Solvers

B-17
B-19
B-20
B-22
B-24
B-25
B-26
B-29
B-30
B-35
B-38
B-44
B-45
B-46

Ken Siler
F . D. P a r k e r
S. L. Basin
Ken Siler
Sr. M a r y De Sales McNabb, C h a r l e s R. Wall
Douglas Lind, C h a r l e s R. Wall
Gurdial Singh
Douglas Lind
Douglas Lind
Denis Hanson
B r i a n Scott
J. L. Brown., J r .
J. L. Brown., J r .
Church, J r . ; Kenneth E. NewClyde A. B r i d g e r ; C. A.
c o m e r ;Brown,
C h a r l e sJ rR.
. ; CWall
h a r l e s R. Wall
J. L. Brown, J r . ; Douglas Lind; C h a r l e s R. Wall
Douglas
J. L.
Lind, C h a r l e s R. Wall
J. L. Brown, J r . ; C. B. A. Peck; C h a r l e s R. Wall
M. N. S. Swamy, C h a r l e s R. Wall
M. N. S. Swamy, C h a r l e s R. Wall
M. N. S, Swamy, C h a r l e s R. Wall
Howard L. Walton
D e r m o t t A. B r e a u l t ; J a m e s E Desmond; John E. H o m e r ,
J r . ; George Ledin, J r . ; C, B. A. P e c k ; Dean B. P r i e s t
D e r m o t t A. B r e a u l t ; J. L. Brown, J r . ; John E. H o m e r , J r ;
George Ledin, J r . ; C. B. A. Peck; Dean B. P r i e s t
J a m e s E. Desmond; George Ledin, J r .
Douglas Lind
J. L. Brown, J r . ; Douglas Lind.

B-47
B-48
B-49
B-50
B-52
B-53
B-54
B-55
B-70
B-71
B-72
B-74
B-75

XXXXXXXXXXXXXXX

B-86

Proposed by Verner E. Hoggatt, Jr., San Jose State College, San Jose,

California

Show that the squares of every third Fibonacci number satisfy

y

17

n+3

y n + 2 " 1 7 y n+1 + y