• Aucun résultat trouvé

THE INFINITY

N/A
N/A
Protected

Academic year: 2022

Partager "THE INFINITY"

Copied!
3
0
0

Texte intégral

(1)

Annales Academie Scientiarum Fennica Series A. I. Mathematica

Volumen t4, 1989, 129-131

ON THE INFINITY OF THE

LONERGAN_HO SACK PRES ENTATION

Markku Niemenmaa and Gerhard

Rosenberger

Introduction

In

1,984 F.D. Lonergan and J. Hosack [3] introduced the following group pre- sentation:

G: (r,z: z3*z3r-r : zs12zzr'-

r).

According

to

the authors there was good reason

to

believe

that

the presentation is

that

of a finite group. However, they reported that attempts to prove finiteness by using the Todd-Coxeter algorithm were unsuccessful.

By using the computer algebra system CAYLEY

M.

Slattery [5] managed to show in 1985

that

G is infinite. To be precise, Slattery showed

that

G has a factor group which is infinite.

In this short note we wish to point out

that

a

lot

more can be said about the structure

of G.

In fact, we shall consider the presentation

G(n)

: larz

:

znazn0-r :

zn*za222x2

:1)

with n ) 2. In

particular, our results hold

for G: G(3).

For the background

material of this note the reader is advised

to

consult [7].

Main theorem

Consider

the

presentation

G(n)

given

in

the introduction

with n ) 1. If n:7,

then

zrz : n2z

and z5

: z-5: 14.

Consequently G(1) is a finite group of order

40.

Now we shall establish the

Theorem. Let n ) 2

and consider the presentation

G(n).

-lfow

(t)

G(rz) is

a

nontrivial free product

with

amilgamation,

(2) G(n)

has

a

subgroup of frnite index mapping onto a free group of rank 2

and

G(n)

has a free subgroup of rank 2,

(S)

G(n)

has a generating pair

{u,u}

such that

the

subgroup

(rk,rr)

is åee

of

rank 2 for a sufficiently large integer k,

(4) G(n)

is SQ-universal (i.e. every countable group is embeddable

in

some factor group

of

G(n)).

doi:10.5186/aasfm.1989.1422

(2)

130

M.

.lfiemen maa and G. Rosenberger

Proof. (1)

If n

is even, then the free product

Za* 22: (u,z

: n4

: ,2 :1)

is an epimorphic image

of G(n).If. n

is odd

(then n ) 3),

we have

D - (t,z:

zn

-

(22x2)'

: 1)

as an epimorphic image

of G(n).

Now

D

is a nontrivial free product of

är

and

Hz withthe

amalgamated subgrouP

ä,

where

H1: (rl = Z,

Hz: (y,r, ," : (r2y)2 - 1) and ä : l*') = fu). By

Lemma 3.2

of

[6]

we conclude

thal G(n)

is a nontrivial free product

with

amalgamation for every

n)

2.

(2) The free product

Za*

22 ha,s the triangle-group

fQ,4,5)-

(o,b i a2

-

b4

-

(

ob)'-

1)

as an epimorphic image and clearly G(n.) has

T(2,4,5)

as an epimorphic im- age provided

n

is even. Next consider

D

(the free product

with

amalgamation) introduced in the

first

part of the proof. Now

D

has the triangle-group

T(n,7,2) :

(o,b : an

-

b7

-

( ob)'

-

1)

as an epimorphic image Thus

G(")

has as

and naturally

G(")

has

T(",7,2)

as an epimorphic image an epimorphic image a triangle-group

T(p,

q,r) :

(a,b ; aP

:

bq

:

(ab)'

:

Ll

with 2 Sp,q,r

and

(tld+Olil+Glr) <

1

.

Now

T(p,q,r)

has a surface group

F(g)

of genus g

2 2

as a subgroup of finite index (see

[7]).

Since

F(g) k 2Z)

has a free group of

rank 2

as an epimorphic image,

it

follows

by

considering the pre-image

ot F(g) in G(n) that G(n)

has a subgroup of finite index which maps onto a free group

ofrank

2 and consequentl1

G(")

has a free subgroup

ofrank

2.

(3) Since

T(p,q,r)

can be regarded as a subgroup of P,S.L(2,.E) (see [7]), our assertion follows from [4].

(a)

BV [1] we know

that

a free group of

rank 2 is

SQ-universal. Clearly, a pre-image of an sQ-universal group is sQ-universal. Finally, by [2]

it

follows that a group which has an SQ-universal subgroup of finite index is also SQ-universal.

The proof is complete.

Remarks.

As indicated in the beginning of this paper G(1) is a finite group of order 40 and consequently

it

has elements of finite order

) 2.

If we consider G(2) , then

z2rzzr-r : z4r2z2a2:

1 implies o2z2

:

zza2

,

Ftrthermorer z6

: r-4and

2-6: r-4,

hence

zr2:L

and

18: L.

Now we state

Problem 1.

Is

it

true

that G(n)

has elements of finite order

) 2 fot n)

3?

We also state

Problem 2.

Is

it

true

that G(n)

has a torsion-free normal subgroup of finite

indexfor n)2?

(3)

On the

infinity

of the Lonergan-Hosack presel2 tation 131 References

tU

HIGMAN, G., B.H. NsuulNN, and H. NpulrlNN: Embedding theorems for groups. - J.

London Math. Soc. 24, t949, 247-254.

l2l

NouMl.NN, P.M.: The SQ-universality of some finitely generated groups.

-

J. Austral.

Math. Soc. L6, 1973, 1-6.

t3]

LoNnRGlN, F.D., and J. Hos,q.cr: Problem 308.- Notices Amer. Math. Soc. 31, 5, 1984.

t4]

RospNspRopR, G.: On discete free subgroups of linea.r groups.- J. London Math. Soc.

(2) 17, 1978, 79-85.

t5]

Sr,nrreRy, M.: Solution to AMS Notices query No. 308. - Cayley Bulletin 2, 1985,37-39.

t6]

ZIoscu.a.uc, H.: On decompositions of discontinuous groups of the plane. - Math. Z. 151, 1976, 165-188.

t7]

ZrESCn.LNc, H., E.

Vocr,

and H.D. Colopwpv: Surfaces a.nd planar discontinuous groups.

-

Lecture Notes in Mathematics 835. Springer-Verlag, Berlin-Heidelberg- New York, 1980.

University of Oulu University of Dortmund

Department of Mathematics Department of Mathematics

SF-90570 Oulu D-4600 Dortmund 50

Finland Federal Republic of Germany

Received 8 February 1988

Références

Documents relatifs

The activity of writing on antiquities is recorded other times in his manuscripts at the third person : “Pyrrho Ligorio che ha scritta quest'opera (taur.13,f.14)” or “Pyrrho che

Cette étude nous montre que la Dimedone et l’ortho-phénylènediamine deux produits commerciaux et bon marché, réagissent pour donner l’intermédiaire

Due to the fact that heuristic algorithms give only approximate results it is unlikely to obtain fitness function equal to 0, however presented results, especially from Ant

We used Maple 14 to solve optimal stopping prob- lem for geometric random walk and managed to find explicit formulas for this problem.. Also for geometric random walk explicit

As a result, the authors have focused on a method which is as efficient as possible, as it motivates students to applied math and to the technology and saves time

On the other hand, for the product of larger matrices (or polynomials), Residue Number Systems often perform better than direct implementations (see Section 4.2), and as such, they

Best for matrix/polynomial size integer bitsize Change of representation of integers RNS conversion is frequently a bottleneck... How to perform arithmetic operations

Best for matrix/polynomial size integer bitsize Conversion to/from Residue Number System is frequently a bottleneck... How to perform arithmetic operations