• 検索結果がありません。

(1)The free ommutative automorphi 2-generated loop of nilpoteny lass 3 Dylene Agda Souza de Barros, Alexander Grishkov, Petr Vojtehovsk y Abstrat.Aloop isautomorphiifallitsinnermappings areautomorphisms

N/A
N/A
Protected

Academic year: 2022

シェア "(1)The free ommutative automorphi 2-generated loop of nilpoteny lass 3 Dylene Agda Souza de Barros, Alexander Grishkov, Petr Vojtehovsk y Abstrat.Aloop isautomorphiifallitsinnermappings areautomorphisms"

Copied!
16
0
0

読み込み中.... (全文を見る)

全文

(1)

The free ommutative automorphi

2-generated loop of nilpoteny lass 3

Dylene Agda Souza de Barros,

Alexander Grishkov, Petr

Vojtehovsk

y

Abstrat.Aloop isautomorphiifallitsinnermappings areautomorphisms.

Weonstrutthefreeommutativeautomorphi2-generatedloopofnilpoteny

lass3. Ithasdimension8overtheintegers.

Keywords:free ommutative automorphi loop, automorphi loop, assoiator

alulus

Classiation: 20N05

1. Introdution

AloopisanonemptysetQwithabinaryoperationsuhthatforeverya2Q

the left and right translations L

a , R

a

: Q ! Q, bL

a

= ab, bR

a

= ba are

bijetionsofQ,andthereisanidentityelement12Qsatisfying1a=a1=a

for all a 2 Q. We will also write the multipliation asjuxtaposition, and we

willusetoindiate thepriorityofmultipliations. Forinstane, a(bd)means

a((b)d).

ThemultipliationgroupofQisthepermutationgroupMlt(Q)=hL

a

;R

a

; a2

Qi generated by all left and right translations. The stabilizer of 1 in Mlt(Q)

is the inner mapping group Inn(Q). It is well known, f. [3℄, that Inn (Q) =

hL

a;b

;R

a;b

; T

a

; a;b 2 Qi, where L

a;b

= L

a L

b L

1

ba , R

a;b

= R

a R

b R

1

ab , T

a

=

R

a L

1

a .

AloopQ isautomorphi ifInn(Q)Aut (Q),that is,ifeveryinner mapping

ofQisanautomorphismofQ. Notethatgroupsareautomorphiloops.

Automorphiloopswererststudiedin[4℄,whereitwasproved,amongother

results, that automorphi loops form a variety and are power-assoiative, that

is, every element generates a group. It was shown in [9℄ that automorphi

loops have the antiautomorphi inverse property (ab) 1

= b 1

a 1

. In parti-

ular,ommutativeautomorphiloopshavetheautomorphiinverse property, or

Dylene Agda Souza de Barros supported by FAPESP - proess number 2010/16112-4.

Alexander Grishkov supported by FAPESP and CNPq (Brazil). The researh stay of Dy-

leneAgdaSouza deBarrosandAlexanderGrishkovattheUniversityofDenverwaspartially

(2)

AIP,(ab) =a b . Foranintrodutionto thestruturaltheoryandthehis-

toryofautomorphiloops,see [11℄. Foranintrodutiontothestruturaltheory

ofommutativeautomorphiloops,see[6℄,[7℄.

Thispaperisonernedwithfreeobjetsinthevarietyof ommutativeauto-

morphiloops,inpartiularwiththefreeommutativeautomorphi2-generated

loopofnilpotenylass3.

Theenter ofaloopQistheassoiativesubloopZ(Q)=fa2Q; a'=afor

every'2Inn (Q)g. ThusZ(Q)onsistsofallelementsa2Qthat ommuteand

assoiatewithallotherelementsofQ. DeneZ

0

(Q)=1,Z

1

(Q)=Z(Q),andfor

i1letZ

i+1

(Q)bethepreimageofZ(Q=Z

i

(Q))undertheanonialprojetion

Q!Q=Z

i

(Q). ThenaloopQisnilpotent oflass nifZ

n 1

(Q)6=Q=Z

n (Q).

Itwasshownindependently in[5℄and[8℄that foranoddprimepeveryom-

mutativeautomorphi loopoforderp k

isnilpotent. By[8℄,aommutativeauto-

morphiloopof orderp 2

is aommutativegroup,butthere exist nonassoiative

ommutativeautomorphi loopsof order p 3

|these were onstruted in [7℄ and

lassieduptoisomorphismin[2℄.

Oneofthemaintoolsusedinthelassiation[2℄wasthedesriptionofthefree

ommutativeautomorphi2-generatedloopofnilpotenylass2. Thispaperan

thereforebeseenasanaturalontinuationoftheprogrambegunin[2℄. Arelated

projetis [10℄, where heavyassoiatoraluluswasused todetermine thebases

andordersoffreeommutativeMoufangloopswithupto sevengenerators.

Forn2,letF

n

(x;y)bethefreeommutativeautomorphiloopofnilpoteny

lassnonfreegeneratorsx,y.

Forelementsa,b,ofaloopQdenoteby(a;b;)theassoiator ofa,b,,that

is,theuniqueelementsatisfyingtheequationab=(ab)(a;b;).

WeobtainedthefollowingdesriptionofF

2

(x;y)in [2℄:

Theorem1.1([2,Theorem2.3℄). LetF

2

(x;y)bethefreeommutativeautomor-

philoopofnilpotenylass2withfreegeneratorsx,y,andletu

1

=(x;x;y),u

2

=

(x;y;y). Theneveryelementof F

2

(x;y)anbewrittenuniquelyasx a

1

y a

2

u a3

1 u

a4

2

forsomea

1

;a

2

;a

3

;a

4

2Z,andthemultipliationinF

2

(x;y)isgivenby

(x a1

y a2

u a3

1 u

a4

2 )(x

b1

y b2

u b3

1 u

b4

2 )=x

a1+b1

y a2+b2

u

a3+b3 a1b1(a2+b2)

1

u

a4+b4+a2b2(a1+b1)

2

:

Aswearegoingtosee,to desribeF

3

(x;y)isonsiderablymorediÆult.

Letusallanassoiatorompounded ifitisoftheform (a;b;)whereatleast

oneofa,b,isagainanassoiator(u;v;w). Itiseasytosee, f.Proposition2.1,

that aommutativeloop isof nilpoteny lassat most3ifand only ifall om-

pounded assoiatorsareentral.

Ultimately we provein Theorem 5.4 that everyelement of F

3

(x;y) is of the

anonialform

(x a

1

y a

2

u a3

u a4

)v a5

v a6

v a7

v a8

;

(3)

where

u

1

=(x;x;y); u

2

=(x;y;y); v

1

=(x;x;u

1 );

v

2

=(x;x;u

2

); v

3

=(y;y;u

1

); v

4

=(y;y;u

2 );

and where themultipliation formula is asin Lemma 5.3. (The anonial form

anbeparsed unequivoally beausethe ompounded assoiatorsv

1 , v

2 , v

3 , v

4

areentral.) Thisisaomplishedin aseriesofsteps:

InSetion2westudysymmetriesand linearpropertiesoftheassoiatormap

(; ; )inommutativeautomorphiloopsofnilpotenylass3. Weonludethat

inF

3

(x;y)itsuÆestolookatompoundedassoiatorsoftheform(a;b;(;d;e))

where eah a, b, , d, e is either x or y. InSetion 3 we study powers within

assoiators and derive a formula for (a i

;b j

; k

). In Setion 4 we disover sev-

eral nontrivialrelationsamong ompounded assoiatorsof F

3

(x;y),reduingall

ompoundedassoiatorstojust v

1 ,v

2 ,v

3 ,v

4 .

ThemultipliationformulaforF

3

(x;y)isderivedinLemma5.3. Aritialstep

inprovingthemainresult,Theorem5.4, onsistsofshowingthatthemultiplia-

tionformulaofLemma 5.3atuallyyieldsanautomorphiloop. Thisfollowsby

straightforwardalulation(onemerelyneedstoverifythatthegeneratorsL

a;b of

theinnermappinggroupareautomorphisms),butthealulationisextremelyte-

diousanderror-proneandwehavethereforedeidedtodelegateittoaomputer.

TheMathematia[12℄odethataomplishesthealulationanbedownloaded

fromthewebsiteofthethird-namedauthor,www.math.du.edu/~petr. Onewe

knowthattheformulaofLemma 5.3yieldsanautomorphiloopQ,itiseasyto

showthatF

3

(x;y)isfreeandisomorphitoQ.

Reallthattheassoiator subloop A(Q)ofQistheleastnormalsubloopofQ

ontainingallassoiators(soQ=A(Q)isagroup). TheleftnuleusN

(Q),middle

nuleus N

(Q) andright nuleus N

(Q) onsist ofall elementsa2Qsuh that

(a;b;) =1, (b;a;) =1, (b;;a)=1 foreveryb, 2Q, respetively. Thenthe

nuleus N(Q) isdened byN(Q)=N

(Q)\N

(Q)\N

(Q). Weonludethe

paperbyalulatingtheassoiatorsubloop,nuleiandtheenterofQ=F

3 (x;y).

Remark 1.2. Inthebeginningofthispapertheproofsweoergiveallthedetails,

butlateronwegraduallyrelymoreandmoreonthereadertoprovideintermediate

stepsinalulations. Allsuhstepsanbeobtainedinastraightforwardfashion,

albeitsometimeswithonsiderabletimeommitment. Moredetailswillbefound

inthedissertation[1℄ oftherst-namedauthor.

2. Symmetryand linearity in assoiators

Reall that the assoiator in any loop Q is well-dened modulo Z(Q), that

is, (a;b;) = (az

1

;bz

2

;z

3

) for any a, b, 2 Q and z

1

;z

2

;z

3

2 Z(Q). In any

ommutativelooptheidentity

(4)

holdsbeauseaba=aab=aba. Itiswellknownthatinanyommutativeloop

ofnilpotenylass2wehave(a;b;)=(;b;a) 1

,(a;b;)(b;;a)(;a;b)=1. We

willuse allthese observationsand thefollowingwell-knownpropositionwithout

referene.

Proposition 2.1. LetQbeaommutativeloop.

(i) Qhasnilpotenylassatmost2ifandonlyifallassoiatorsareentral.

(ii) Qhasnilpotenylassatmost3ifandonlyifallompoundedassoiators

((a;b;);d;e),(a;(b;;d);e),(a;b;(;d;e))areentral.

Proof: Suppose that Q has nilpoteny lass at most 2. Then Q=Z(Q) is an

abelian group. Sine A(Q) is the least normal subloop S suh that Q=S is an

abeliangroup,itfollowsthatA(Q)Z(Q). Theonverseisprovedbyreversing

theargument.

Let us write a for aZ(Q) 2 Q=Z(Q). Suppose that Q has nilpoteny lass

at most 3. Then Q=Z(Q) has nilpoteny lass at most 2 and thus (a;b;) 2

Z(Q=Z(Q)) for every a;b; 2 Q by (i). This is equivalent to ((a ;b;);d ;e) =

(d ;(a;b;);e)=(d;e;(a;b;))=1

Q=Z(Q)

andthusto((a;b;);d;e),(d;(a;b;);e),

(d;e;(a;b;))2Z(Q). Theonverseis againproved by reversingtheargument.

We proeed to show that in a ommutative automorphi loop of nilpoteny

lass2theassoiatorislinearin alloordinates.

Lemma 2.2. Let Qbealoopand leta;b;2 Qbesuh that (a;b;) 2Z(Q).

ThenL

b;a

=(a;b;) 1

,aR

b;

=a(a;b;), andbL

a R

L

1

a R

1

=b(a;b;).

Proof: Sine(a;b;)isentral,wehaveab(a;b;) 1

=(ab)(a;b;) 1

=ab,

orL

b L

a

=((a;b;) 1

)L

ab ,orL

b;a

=(a;b;) 1

. Also,ab=a(a;b;)b,or

aR

b R

=(a(a;b;))R

b ,oraR

b;

=a(a;b;). Finally,ab=a(b(a;b;))yields

thelastequality.

Proposition 2.3. Let Q be an automorphi loop of nilpoteny lass 2. Then

(ab;;d) =(a;;d)(b;;d),(a;b;d) =(a;b;d)(a;;d), (a;b;d)=(a;b;)(a;b;d)

foreverya;b;;d2Q.

Proof: Sine Q is automorphi, the inner mapping R

;d

is an automorphism.

By Lemma 2.2, ab(ab;;d) = (ab)R

;d

= aR

;d bR

;d

= a(a;;d)b(b;;d) =

ab(a;;d)(b;;d)and(ab;;d)=(a;;d)(b;;d)follows. Consequently,(a;b;d)=

(d;b;a) 1

=((;b;a)(d;b;a)) 1

=(;b;a) 1

(d;b;a) 1

=(a;b;)(a;b;d). Finally,

1L

a R

d L

1

a R

1

d

=1showsthatL

a R

d L

1

a R

1

d

isalsoaninnermapping,soLemma

2.2 implies b(a;b;d) =(b)L

a R

d L

1

a R

1

d

= bL

a R

d L

1

a R

1

d L

a R

d L

1

a R

1

d

=

b(a;b;d)(a;;d)=b(a;b;d)(a;;d),andwearedoneuponanelingb.

HereisaloalversionofProposition2.3:

Lemma 2.4. Let Q be an automorphi loop of nilpoteny lass 3, and let

(5)

(i) If(a;;d),(b;;d)2Z(Q)then(ab;;d)=(a;;d)(b;;d).

(ii) If(a;b;d),(a;;d)2Z(Q)then(a;b;d)=(a;b;d)(a;;d).

(iii) If(a;b;), (a;b;d)2Z(Q) then(a;b;d)=(a;b;)(a;b;d).

Proof: Let us prove (ii). Sine Q=Z(Q) is automorphi of nilpoteny lass 2,

Proposition 2.3 implies (a;b;d) = (a;b;d)(a;;d)z for some z 2 Z(Q). This

meansthat(a;b;d) isentral,too. Thenthealulationattheendoftheproof

ofProposition2.3isstillvalid(sineallassoiatorsinvolvedinthealulationare

entral). Theproofsfor(i)and(iii)aresimilar.

Lemma 2.5. Let Q be aommutativeautomorphi loop of nilpoteny lass 3.

Then

((a;b;);d;e) 1

=(e;d;(a;b;));

(2.2)

(a;(b;;d);e)=(a;e;(b;;d))((b;;d);a;e):

(2.3)

foreverya;b;;d;e2Q.

Proof: Note that z = ((a;b;);d;e) 2 Z(Q) beause Q has nilpoteny lass

3. The identity (a;b;)de = ((a;b;)de)z hene yields ((a;b;)de)z 1

=

(a;b;)de, and we get ((a;b;)de)z 1

= (a;b;)de = ed(a;b;) = (e

d(a;b;))(e;d;(a;b;))=((a;b;)de)(e;d;(a;b;)),whihimplies(2.2). Weal-

ulate

(ae(b;;d))(a;e;(b;;d)) 1

=ae(b;;d)=a(b;;d)e

=(a(b;;d)e)(a;(b;;d);e) 1

=((b;;d)ae)((b;;d);a;e)(a;(b;;d);e) 1

=(ae(b;;d))((b;;d);a;e)(a;(b;;d);e) 1

;

whih implies(a;e;(b;;d)) 1

=((b;;d);a;e)(a;(b;;d);e) 1

,or(2.3).

Lemma 2.6. Let Q be aommutativeautomorphi loop of nilpoteny lass 3.

Then

(2.4) (a;(b;;d);(e;f;g))=((a;b;);d;(e;f;g))=((a;b;);(d;e;f);g)=1

foreverya;b;;d;e;f;g2Q.

Proof: SineQ=Z(Q)isofnilpotenylass2,wehave(efg)(efg) 1

=(e;f;g)z

forsomez2Z(Q). Then

(a;(b;;d);(e;f;g))=(a;(b;;d);(efg)(efg) 1

):

The automorphi inverse property and Lemma 2.4 yield (a;(b;;d);(e;f;g)) =

1. The identity ((a;b;);(d;e;f);g) = 1 follows by (2.2). The argument for

((a;b;);d;(e;f;g))=1issimilar.

(6)

Lemma 2.7. Let Q be aommutativeautomorphi loop of nilpoteny lass 3.

Then

(a;b;)=(;b;a) 1

; (2.5)

(a;b;)=(a;;b)(b;a;) (2.6)

foreverya;b;2Q.

Proof: Wehaveab=(ab)(a;b;)=(ba)(a;b;)=(ba)(;b;a)(a;b;)=

(ab)(;b;a)(a;b;),andthelasttermanberewrittenas(ab)(;b;a)(a;b;)

by(2.4),so(2.5)follows. Similarly,ab=(ab)(a;b;)=(ba)(b;;a)(a;b;)=

((ab)(;a;b)(b;;a))(a;b;),thelasttermequals(ab)((;a;b)(b;;a)(a;b;))

by(2.4)andLemma 2.4,sowehave(;a;b)(b;;a)(a;b;)=1. ByLemma2.6,

theAIP and (2.5), weget(a;b;)= ((;a;b)(b;;a)) 1

=(;a;b) 1

(b;;a) 1

=

(b;a;)(a;;b),whihis(2.6).

Notethatin anyommutativeloopofnilpotenylass3wehave

(2.7) aL

b;

=a(a;b;)(b;a;(a;b;));

beause

aL

b;

=(ba)L 1

b

=(ab)L 1

b

=((ab)(a;b;))L 1

b

=((ba)(a;b;))L 1

b

=(ba(a;b;)(b;a;(a;b;)))L 1

b

=a(a;b;)(b;a;(a;b;)):

Proposition 2.8. Let Q be a ommutative automorphi loop of nilpoteny

lass3. Then

(ab;;d)=(a;;d)(b;;d)((a;;d);a;b)((b;;d);b;a)

((a;;d);b;)((b;;d);a;)((a;;d);b;d)((b;;d);a;d);

(a;b;d)=(a;b;)(a;b;d)((a;b;);;d)((a;b;d);d;)

((a;b;);d;b)((a;b;d);;b)((a;b;);d;a)((a;b;d);;a);

(a;b;d)=(a;b;d)(a;;d)((a;b;d);b;)((a;;d);;b)

((a;b;d);;a)((a;;d);b;a)((a;b;d);;d)((a;;d);b;d):

Proof: By(2.7),theidentity(ab)L

;d

=aL

;d bL

;d

anberewrittenas

ab(ab;;d)(d;ab;(ab;;d))=a(a;;d)(d;a;(a;;d))b(b;;d)(d;b;(b;;d))

=(a(a;;d)b(b;;d))(d;a;(a;;d))(d;b;(b;;d)):

ByLemma 2.4and(2.4),(2.5),wehave

a(a;;d)b(b;;d)=(a(a;;d)b)(b;;d)(a(a;;d);b;(b;;d)) 1

=(a(a;;d)b)(b;;d)(a;b;(b;;d)) 1

((a;;d);b;(b;;d)) 1

(7)

=(ba(a;;d))(b;;d)(b;a;(a;;d)) 1

((b;;d);b;a)

=(ba(a;;d))(b;;d)((a;;d);a;b)((b;;d);b;a):

Sine(ab;;d)=(a;;d)(b;;d)zforsomez2Z(Q),Lemma2.4yields

(d;ab;(ab;;d))=(d;ab;(a;;d)(b;;d))=(d;ab;(a;;d))(d;ab;(b;;d))

=(d;a;(a;;d))(d;b;(a;;d))(d;a;(b;;d))(d;b;(b;;d)):

Uponsubstituting andanelingab and likeassoiators,theidentity (ab)L

;d

=

aL

;d bL

;d

thereforebeomes

(ab;;d)(d;b;(a;;d))(d;a;(b;;d))=(a;;d)(b;;d)((a;;d);a;b)((b;;d);b;a):

Theformulafor(ab;;d) nowfollowsbyLemma2.4and(2.5).

Note that Lemma 2.4 and (2.5) imply ((a;b;);d;e) 1

= ((a;b;) 1

;d;e) =

((;b;a);d;e). This observation and (2.5) applied to the formula for (ab;;d)

yieldtheformulafor(a;b;d).

Using(2.6),wealulate

(a;b;d)=(a;d;b)(b;a;d)=(a;d;b)(a;d;)(b;a;d)(;a;d)

((a;d;b);b;)((a;d;);;b)((a;d;b);;d)((a;d;);b;d)((a;d;b);;a)((a;d;);b;a)

((b;a;d);b;)((;a;d);;b)((b;a;d);;a)((;a;d);b;a)((b;a;d);;d)((;a;d);b;d):

Therstfourassoiatorsassoiateby(2.4),so(a;d;b)(a;d;)(b;a;d)(;a;d) =

(a;d;b)(b;a;d)(a;d;)(;a;d)=(a;b;d)(a;;d). Weansimilarlypairtheom-

poundedassoiators,usingLemma2.4. Forinstane,((a;d;b);b;)((b;a;d);b;)=

((a;d;b)(b;a;d);b;)=((a;b;d);b;). Theformulafor(a;b;d) follows.

We an now deal with produts in all arguments of a ompounded assoia-

tor. Forprodutsoftheform (ab;;(d;e;f))weanuseLemma2.4(orProposi-

tion2.8),andforproduts(a;b;(d;e;f))wenotethat(d;e;f)=(;e;f)(d;e;f)z

for someentral elementz (the expliitform ofz follows from Proposition 2.8)

and alulate(a;b;(;e;f)(d;e;f))=(a;b;(;e;f))(a;b;(d;e;f))byLemma 2.4.

Fromnowon,wewillusetheseandsimilaridentities,oftenwithoutexpliitrefe-

rene.

3. Powerswithin assoiators

UsingProposition2.8,weproeedtoderiveformulaeforpowerswithinassoi-

ators. Dene:Z!Z,:Z!Zby

(3.1) (n)=(n

3

n)=3; (n)=n 2

n:

Note that (0) =(0) =0, (n+1) =(n)+n 2

+n, (n+1) =(n)+2n,

2

(8)

Lemma 3.1. Let Q be aommutativeautomorphi loop of nilpoteny lass 3.

Then

(a n

;b;)=(a;b;) n

((a;b;);a;a) (n)

((a;b;);a;b) (n)

((a;b;);a;) (n)

; (3.2)

(a;b n

;)=(a;b;) n

((a;b;);b;b) (n)

((a;b;);b;a) (n)

((a;b;);b;) (n)

; (3.3)

(a;b;

n

)=(a;b;) n

((a;b;);;) (n)

((a;b;);;a) (n)

((a;b;);;b) (n)

(3.4)

foreverya;b;2Qandeveryn2Z.

Proof: We prove (3.2); the equations (3.3), (3.4) are proven analogously. If

n = 0, (3.2) holds. Suppose that (3.2) holds for some n 0. Note that

((a i

;b;);a j

;d) = ((a;b;) i

;a j

;d) =((a;b;);a;d) ij

for everyi, j 0, by Lem-

ma2.4,usingourusualtrik(a i

;b;)=(a;b;) i

z forsomez2Z(Q). ByPropo-

sition2.8wethenhave

(a n+1

;b;)=(aa n

;b;)

=(a;b;)(a n

;b;)((a;b;);a;a n

)((a n

;b;);a n

;a)

((a;b;);a n

;b)((a n

;b;);a;b)((a;b;);a n

;)((a n

;b;);a;)

=(a;b;)(a n

;b;)((a;b;);a;a) n

2

+n

((a;b;);a;b) 2n

((a;b;);a;) 2n

=(a;b;) n+1

((a;b;);a;a) (n)+n

2

+n

((a;b;);a;b) (n)+2n

((a;b;);a;) (n)+2n

=(a;b;) n+1

((a;b;);a;a) (n+1)

((a;b;);a;b) (n+1)

((a;b;);a;) (n+1)

:

Asfornegativepowers,rstnotethat Proposition2.8gives

1=(1;b;)=(aa 1

;b;)=(a;b;)(a 1

;b;)((a;b;);a;a 1

)((a 1

;b;);a 1

;a)

((a;b;);a 1

;b)((a 1

;b;);a;b)((a;b;);a 1

;)((a 1

;b;);a;)

=(a;b;)(a 1

;b;)((a;b;);a;b) 2

((a;b;);a;) 2

:

Sineassoiatorsassoiatewithoneanotherby(2.4),wededue

(a 1

;b;)=(a;b;) 1

((a;b;);a;b) 2

((a;b;);a;) 2

:

Thenforeveryn>0wehave

(a n

;b;)=((a n

) 1

;b;)=(a n

;b;) 1

((a n

;b;);a n

;b) 2

((a n

;b;);a n

;) 2

=(a;b;) n

((a;b;);a;a) (n)

((a;b;);a;b) 2n

2

(n)

((a;b;);a;) 2n

2

(n)

=(a;b;) n

((a;b;);a;a) ( n)

((a;b;);a;b) ( n)

((a;b;);a;) ( n)

;

nishingtheproofof (3.2).

Lemma 3.2. Let Q be aommutativeautomorphi loop of nilpoteny lass 3.

Then

i j k ijk (i)jk (i)j

2

k (i)jk

2

(9)

((a;b;);b;a) i(j)k

((a;b;);b;b) i(j)k

((a;b;);b;) i(j)k

2

((a;b;);;a) ij(k )

((a;b;);;b) ij(k )

((a;b;);;) ij(k )

foreverya;b;2Qandi;j;k2Z.

Proof: ByLemmas2.4, 3.1andProposition2.8, (a i

;b j

; k

)isequalto

(a;b j

; k

) i

((a;b j

; k

);a;a) (i)

((a;b j

; k

);a;b j

) (i)

((a;b j

; k

)a;

k

) (i)

=(a;b j

; k

) i

((a;b;);a;a) (i)jk

((a;b;);a;b) (i)j

2

k

((a;b;);a;) (i)jk

2

;

theterm(a;b j

; k

) i

isequalto

[(a;b;

k

) j

((a;b;

k

);b;b) (j)

((a;b;

k

);b;a) (j)

((a;b;

k

);b;

k

) (j)

℄ i

=(a;b;

k

) ij

((a;b;);b;b) i(j)k

((a;b;);b;a) i(j)k

((a;b;);b;) i(j)k

2

;

andtheterm(a;b;

k

) ij

isequalto

[(a;b;) k

((a;b;);;) (k )

((a;b;);;a) (k )

((a;b;);;b) (k )

℄ ij

=(a;b;) ijk

((a;b;);;) ij(k )

((a;b;);;a) ij(k )

((a;b;);;b) ij(k )

:

4. Redution

SineweultimatelywanttodesribethefreeloopF

3

(x;y),wewillfromnowon

startfousingonformulaethatinvolveonlytwovariablesx,y. Forxedelements

x,y (not neessarilythegeneratorsofF

3

(x;y)),let

u

1

=(x;x;y); u

2

=(x;y;y);

z

1

=(x;x;u

1

); z

2

=(x;x;u

2

); z

3

=(x;y;u

1

); z

4

=(x;y;u

2 );

z

5

=(y;x;u

1

); z

6

=(y;x;u

2

); z

7

=(y;y;u

1

); z

8

=(y;y;u

2 ):

Noadditional assoiatorswill be needed sine(x;y;x) =(y;x;y)= 1by (2.1),

ompoundedassoiatorsoftheform((; ; ); ; )and(;(; ; ); )anberewritten

asprodutsofompoundedassoiatorsoftheform( ; ;(; ; ))by(2.2)and(2.3),

ompoundedassoiatorswithassoiatorsintwoomponentsvanishby(2.4),and

produtswithinassoiatorsanbehandled byProposition2.8.

Thereadermighthavenotiedthatinourprodutformulas(suhasinPropo-

sition 2.8) we aumulate assoiators on the left, but we hose the anonial

ompounded assoiatorswithaumulatedassoiatorsontheright. It iseasyto

onvertbetweenthese twoformats,sine

1 1 1

(10)

by(2.5)andLemma 2.4. Alsonote that

(a;b;(;d;e))=(a;b;(e;d;) 1

)=(a;b;(e;d;)) 1

thanksto (2.5)and Lemma2.4.

Lemma 4.1. Let Q be aommutativeautomorphi loop of nilpoteny lass 3.

Thenforeveryx;y2Qwehavez

2

=z

3

=z

5 andz

4

=z

6

=z

7 .

Proof: Fousingrstontheprodutinthethirdoordinate,wealulate,start-

ingwith(2.1):

1=(xy;x;xy)=(xy;x;x)(xy;x;y)((xy;x;x);x;y)((xy;x;y);y;x)

((xy;x;x);y;x)((xy;x;y);x;x)((xy;x;x);y;xy)((xy;x;y);x;x y)

=(xy;x;x)(xy;x;y)z

5 z

1

3 z

3 z

1

1 z

3 z

7 z

1

1 z

1

5

=(xy;x;x)(xy;x;y)z 2

1 z

3 z

7

=(y;x;x)((y;x;x);y;x)((y;x;x);x;x)((y;x;x);x ;x)

(x;x;y)((x;x;y);x;y)((x;x;y);y;x)((x;x;y);y;y)z 2

1 z

3 z

7

=(y;x;x)(x;x;y)z

3 z

1 z

1 z

1

5 z

1

3 z

1

7 z

2

1 z

3 z

7

=1z

3 z

1

5 :

Henez

3

=z

5

,or(x;y;(x;x;y))=(y;x;(x;x;y)). Interhangingx andy in this

identityyields(y;x;(y;y;x))=(x;y;(y;y;x)),whihisequivalenttoz

6

=z

4 .

In the following alulation we will also use (2.7) and Lemma 3.1. As Q is

automorphi,yL

x;x (xy)L

x;x

=(yxy)L

x;x

. Onthelefthandsideofthisidentity

wehave

yL

x;x

=y(y;x;x)(x 2

;y;(y;x;x))=y(y;x;x)z 2

3

and

(xy)L

x;x

=(xy)(xy;x;x)(x 2

;xy;(xy;x;x))=(xy)(xy;x;x)(x 2

;xy;(y;x;x))

=(xy)(y;x;x)((y;x;x);y;x)((y;x;x);x;x) 2

z 2

1 z

2

3

=(xy)(y;x;x)z

3 z

2

1 z

2

1 z

2

3

=(xy)(y;x;x)z 1

3

;

whileontherighthand sidewehave

(yxy)L

x;x

=(yxy)(yxy;x;x)(x 2

;yxy;(yxy;x;x))

=(yxy)(y 2

x(y;y;x) 1

;x;x)(x 2

;y 2

x;(y 2

;x;x))

=(yxy)(y 2

x;x;x)((x;y;y);x;x)z 8

3 z

4

1

=(xy 2

(x;y;y))(xy 2

;x;x)z 1

2 z

8

3 z

4

1

=(xy 2

(x;y;y))(y 2

;x;x)((y 2

;x;x);y 2

;x)((y 2

;x;x);x;x) 2

z 1

2 z

8

3 z

4

1

=(xy 2

(x;y;y))(y;x;x) 2

((y;x;x);y;y) 2

((y;x;x);y;x) 4

z 4

3 z

4

1 z

1

2 z

8

3 z

4

1

=xy 2

(x;y;y)(y;x;x) 2

z 2

z 4

z 4

z 1

=xy 2

(x;y;y)(y;x;x) 2

z 2

z 1

:

(11)

Returningtothelefthandside,werewriteitas

y(y;x;x)(xy)(y;x;x)z 3

3

=(y(y;x;x)xy)(y;x;x)(y(y;x;x);xy;(y;x;x)) 1

z 3

3

=(xyy(y;x;x))(y;x;x)(y;xy;(y;x;x)) 1

z 3

3

=(xyy)(y;x;x)(xy;y;(y;x;x)) 1

(y;x;x)z

5 z

7 z

3

3

=(xyy)(y;x;x) 2

z

3 z

7 z

5 z

7 z

3

3

=(xy 2

)(x;y;y)(y;x;x) 2

z 2

3 z

5 z

2

7

=xy 2

(x;y;y)(y;x;x) 2

z 2

3 z

5 z

2

7 :

Comparingthe twosidesnowyieldsz 1

2

=z 2

3 z

5 . Butz

2

3 z

5

=z 1

3

bytherst

partofthislemma,andhenez

2

=z

3

. Swithingxandy intheidentityz

2

=z

3

givesz

7

=z

6

.

WiththeredutionofLemma4.1inmind, wesetforanyxedx,y

u

1

=(x;x;y); u

2

=(x;y;y); v

1

=(x;x;u

1 );

v

2

=(x;x;u

2

); v

3

=(y;y;u

1

); v

4

=(y;y;u

2 ):

WearenowreadytodesribeanonialelementsofthefreeloopF

3 (x;y).

Lemma4.2. EveryelementofF

3

(x;y)an bewrittenin theanonialform

(x a1

y a2

u a

3

1 u

a

4

2 )v

a

5

1 v

a

6

2 v

a

7

3 v

a

8

4

;

wherea

i 2Z.

Proof: Let X = fx;x 1

;y;y 1

g. We rst note that any assoiator an be

written as u b

1

1 u

b

2

2 Q

v

i

i

. Indeed, sine F

3

(x;y) has nilpoteny lass three, no

ompounded assoiatorsappear within assoiators. Using Lemma 2.4, Proposi-

tion2.8and theironsequenes,everyassoiatoranbewritten asaprodutof

ompoundedassoiatorsandordinaryassoiatorswithallvariablesinX. Infat,

equations(2.4),(2.1),(2.2)and(2.3)imply thateveryassoiatorisaprodutof

u

1 , u

2 , the z

i

sand theirinverses. Sine assoiatorsassoiateamong themselves

by(2.4),thisprodutisoftheformu b

1

1 u

b

2

2 Q

z d

i

i

forsuitableexponentsinZ,and

heneoftheform u b1

1 u

b2

2 Q

v i

i

byLemma4.1.

To establish the lemma, it suÆes to show that a produt of two anonial

words is also anonial. First, [x a1

y a2

u a3

1 u

a4

2 Q

v i

i

℄[x b1

y b2

u b3

1 u

b4

2 Q

v di

i

℄ =

(x a1

y a2

u a

3

1 u

a

4

2 )(x

b1

y b2

u b

3

1 u

b

4

2 )

Q

v

i +d

i

i

,soitsuÆestoshowthattheprodut

(x a

1

y a

2

u a3

1 u

a4

2 )(x

b

1

y b

2

u b3

1 u

b4

2

)hasthedesiredform. Weanrewritethisword

as ((x a1

y a2

x b1

y b2

)u a

3

1 u

a

4

2 )u

b

3

1 u

b

4

2

w with some produt w of ompounded

assoiators,andfurtherto(x a1

y a2

x b1

y b2

)u a3+b3

1 u

a4+b4

2

w,usingLemma2.4and

(2.4). Now,x a

1

y a

2

x b

1

y b

2

anbewrittenas(((x a

1 +b

1

y a

2 +b

2

)t

1 )t

2 )t

k ,where

eaht

i

isanassoiator. UsingLemma2.4and(2.4)again,wefurtherrewritethis

as(x a1+b1

y a2+b2

)(t t t ). Therest iseasy.

(12)

5. The mainresult

The alulation desribed in the proof of Lemma 4.1 is straightforward but

rathertedious. Beforeweattemptit,wenote:

Lemma5.1. IntheloopF

3

(x;y)allassoiatorsarein themiddle nuleus.

Proof: Thanksto(2.1), (2.5)andLemmas2.4, 4.2and Proposition 2.8,itsuf-

es to show that (x;u

1

;y) = (x;u

2

;y) = 1. By (2.3) and (2.5), (x;u

1

;y) =

(x;y;u

1 )(u

1

;x;y) = (x;y;u

1

)(y;x;u

1 )

1

= z

3 z

1

5

= 1, and also (x;u

2

;y) =

(x;y;u

2 )(u

2

;x;y)=(x;y;u

2

)(y;x;u

2 )

1

=z

4 z

1

6

=1,whereweusedLemma4.1.

Reallthemappings, of (3.1).

Lemma5.2. InF

3

(x;y)wehaveforeverya

1

;a

2

;b

1

;b

2 2Z

x a

1

y a

2

x b

1

y b

2

=x a

1 +b

1

y a

2 +b

2

u

a1b1(a2+b2)

1

u

a2b2(a1+b1)

2

v

(a2+b2)(b1(a1)+a1(b1))+a2(a1(b1)+b 2

1

(a1))+b2(b1(a1)+a 2

1 (b1))

1

v 2a

1 a

2 b

1 b

2 (a

1 +b

1 )+(a

2 +b

2 )(a

1 (b

1 )+b

1 (a

1 ))+((a

2 )+(b

2 ))(a

1 b

2

1 +b

1 a

2

1 ) a

2 b

2 (a

1 +b

1 )

2

v

2a1a2b1b2(a2+b2) (a1+b1)(a2(b2)+b2(a2)) ((a1)+(b1))(a2b 2

2 +b2a

2

2

)+a1b1(a2+b2)

3

v (a

1 +b

1 )(a

2 (b

2 )+b

2 (a

2 )) a

1 (a

2 (b

2 )+b

2

2 (a

2 )) b

1 (b

2 (a

2 )+a

2

2 (b

2 ))

4

:

Proof: Using(2.5),wealulate

x a1

y a2

x b1

y b2

=(x a1

y a2

x b1

)y b2

(y b2

;x b1

;x a1

y a2

)

=((x b

1 +a

1

y a

2

)(y a

2

;x a

1

;x b

1

)y b

2

)(y b

2

;x b

1

;x a

1

y a

2

)

=(x a1+b1

y a2

(y a2

;x a1

;x b1

)y b2

)(x a1+b1

y a2

;(y a2

;x a1

;x b1

);y b2

)(y b2

;x b1

;x a1

y a2

):

ByLemma 5.1,weanignoretheompoundedassoiatorandontinue

[(x a

1 +b

1

y a

2

y b

2

)(y a

2

;x a

1

;x b

1

)℄((y a

2

;x a

1

;x b

1

);y b

2

;x a

1 +b

1

y a

2

)(y b

2

;x b

1

;x a

1

y a

2

)

=[(x a1+b1

y a2+b2

(x a1+b1

;y a2

;y b2

))(y a2

;x a1

;x b1

)℄

((y a2

;x a1

;x b1

);y b2

;x a1+b1

y a2

)(y b2

;x b1

;x a1

y a2

):

Beauseassoiatorsassoiatewithoneanother,weanrewritetheformulaas

x a

1

y a

2

x b

1

y b

2

=x a

1 +b

1

y a

2 +b

2

(x a

1 +b

1

;y a

2

;y b

2

)(y a

2

;x a

1

;x b

1

)(y b

2

;x b

1

;x a

1

y a

2

)

((y a2

;x a1

;x b1

);y b2

;x a1+b1

y a2

):

Now,usingLemmas2.4and 4.1freely,

((y a

2

;x a

1

;x b

1

);y b

2

;x a

1 +b

1

y a

2

)=v

a1b1a2b2(a1+b1)

v a

1 b

1 a

2

2 b

2

:

(13)

ByLemma 3.2,

(x a1+b1

;y a2

;y b2

)=u

(a1+b1)a2b2

2

v

(a1+b1)a2b2

2

v

(a1+b1)a 2

2 b2

3

v

(a1+b1)a2b 2

2

3

v

(a1+b1)(a2)b2

3

v

(a1+b1)(a2)b2

4

v (a

1 +b

1 )(a

2 )b

2

2

4

v (a

1 +b

1 )a

2 (b

2 )

3

v (a

1 +b

1 )a

2 (b

2 )

4

v (a

1 +b

1 )a

2 (b

2 )

4

;

and,similarly,

(y a2

;x a1

;x b1

)=u a1b1a2

1 v

(a2)a1b1

3

v (a2)a

2

1 b1

2

v (a2)a1b

2

1

2

v a

2 (a

1 )b

1

2

v a

2 (a

1 )b

1

1

v a

2 (a

1 )b

2

1

1

v a

2 a

1 (b

1 )

2

v a

2 a

1 (b

1 )

1

v a

2 a

1 (b

1 )

1

:

Finally, byProposition2.8and(2.3),weseethat

(y b

2

;x b

1

;x a

1

y a

2

)=(y b

2

;x b

1

;x a

1

)v a

2

1 b

1 a

2 b

2

2 v

a

1 b

2

1 a

2 b

2

2 v

a

1 b

1 a

2 b

2

2

3 :

Theassoiator(y b

2

;x b

1

;x a

1

)anbeobtainedfrom thealreadyalulatedassoi-

ator(y a

2

;x a

1

;x b

1

). Puttingalltheseassoiatorstogether, wearriveat

x a1

y a2

x b1

y b2

=x a1+b1

y a2+b2

u a

1 b

1 (a

2 +b

2 )

1

u a

2 b

2 (a

1 +b

1 )

2

v

1

1 v

2

2 v

3

3 v

4

4

;

where,after summinguptheexponentsoftherespetivev

i

sandsimplifying,

1

=(a

2 +b

2 )(b

1 (a

1 )+a

1 (b

1 ))+a

2 (a

1 (b

1 )+b

2

1 (a

1 ))

+b

2 (b

1 (a

1 )+a

2

1 (b

1 ));

2

=2a

1 a

2 b

1 b

2 (a

1 +b

1 )+(a

2 +b

2 )(a

1 (b

1 )+b

1 (a

1 ))

+((a

2 )+(b

2 ))(a

1 b

2

1 +b

1 a

2

1 ) a

2 b

2 (a

1 +b

1 );

3

= (a

1 +b

1 )(a

2

2 b

2 +a

2 b

2

2 ) (a

1 +b

1 )((a

2 )b

2 +a

2 (b

2 ))

+a

1 b

1 ((a

2 )+(b

2 ))+a

1 b

1 a

2 b

2 (a

2 +b

2 );

4

= (a

1 +b

1 )(a

2 )b

2 (a

1 +b

1 )(a

2 )b

2

2

(a

1 +b

1 )a

2 (b

2 ) (a

1 +b

1 )a

2 (b

2 ):

Theexponents

1 ,

2

alreadyhavethedesiredform. Tomath theexponents

3 ,

4

withtheformulaof thelemma, note that (a+b)=(a)+(b)+ab(a+b)

whilerewriting

3

,andsubstitute(a)=a 2

ainto

4

.

Lemma5.3. InF

3

(x;y)wehave

((x a

1

y a

2

u a3

1 u

a4

2 )v

a5

1 v

a6

2 v

a7

3 v

a8

4 )((x

b

1

y b

2

u b3

1 u

b4

2 )v

b5

1 v

b6

2 v

b7

3 v

b8

4 )

=(x a1+b1

y a2+b2

u a

3 +b

3 a

1 b

1 (a

2 +b

2 )

1

u a

4 +b

4 +a

2 b

2 (a

1 +b

1 )

2

)

v

a5+b5+(a2+b2)(b1(a1)+a1(b1))+a2(a1(b1)+b 2

1

(a1))+b2(b1(a1)+a 2

1

(b1)) a1b1(a3+b3)

1

v a

6 +b

6 +2a

1 a

2 b

1 b

2 (a

1 +b

1 )+(a

2 +b

2 )(a

1 (b

1 )+b

1 (a

1 ))+((a

2 )+(b

2 ))(a

1 b

2

1 +b

1 a

2

1 )

(14)

2 2 1 1 1 1 4 4 3 3 1 2 2 1

v

a7+b7 2a1a2b1b2(a2+b2) (a1+b1)(a2(b2)+b2(a2)) ((a1)+(b1))(a2b 2

2 +b2a

2

2 )

3

+a

1 b

1 (a

2 +b

2 ) a

2 b

2 (a

3 +b

3 ) (a

4 +b

4 )(a

1 b

2 +a

2 b

1 )

v

a8+b8 (a1+b1)(a2(b2)+b2(a2)) a1(a2(b2)+b 2

2

(a2)) b1(b2(a2)+a 2

2

(b2)) a2b2(a4+b4)

4

foreverya

i

;b

i 2Z.

Proof: UsingLemma5.1in therststepand(2.4)intheseond,wehave

(x a

1

y a

2

u a

3

1 u

a

4

2 )(x

b

1

y b

2

u b

3

1 u

b

4

2 )

=x a1

y a2

(u a3

1 u

a4

2 (x

b1

y b2

u b3

1 u

b4

2 ))

=x a1

y a2

(x b1

y b2

u a

3 +b

3

1 u

a

4 +b

4

2 )

=(x a1

y a2

x b1

y b2

)u a

3 +b

3

1 u

a

4 +b

4

2 (x

a1

y a2

;x b1

y b2

;u a

3 +b

3

1 u

a

4 +b

4

2 )

1

:

Nownotethat Lemma2.4yields

(x a

y b

;x

y d

;u e

1 u

f

2 )=v

ae

1 v

af+ade+be

2

v

adf+bf+bde

3

v bdf

3 :

WearethereforedonebyLemma5.2.

IntheproofofthemaintheoremwewilluseaMathematia[12℄odetoverify

ertainproperties ofthemultipliation formulaofLemma 5.3. Theodeanbe

downloadedfromthewebsiteof thethird-namedauthor.

Theorem5.4. LetF

3

(x;y)bethefreeommutativeautomorphiloopofnilpo-

teny lass 3 on free generators x, y. Let u

1

= (x;x;y), u

2

= (x;y;y), v

1

=

(x;x;u

1 ), v

2

= (x;x;u

2 ), v

3

=(y;y;u

1 ), v

4

= (y;y;u

2

). Then eah element of

F

3

(x;y)anbewrittenuniquelyas(x a1

y a2

u a3

1 u

a4

2 )v

a5

1 v

a6

2 v

a7

3 v

a8

4

,andF

3 (x;y)is

isomorphito(Z 8

;),wherethemultipliationofexponentsisasinLemma5.3.

Proof: Let F be dened on Z 8

with multipliation aording to Lemma 5.3.

Denote by e

i

the element of Z 8

whose only non-zero oordinate is equal to 1

and is loated in position i. Straightforwardalulation in Mathematia shows

thatF is aloopwith identityelement(0;0;0;0;0;0;0;0)suhthat (e

1

;e

1

;e

2 )=

e

3 , (e

1

;e

2

;e

2 ) = e

4 , (e

1

;e

1

;e

3 ) = e

5 , (e

1

;e

1

;e

4 ) = e

6 , (e

2

;e

2

;e

3 ) = e

7 and

(e

2

;e

2

;e

4 ) = e

8

. Moreover, F is a ommutative automorphi loop. (Toverify

that F isautomorphi, the ode merelyneeds to hekby symboli alulation

thattheinnermappings L

a;b

areautomorphismsofF.)

Welaim that F

3

(x;y) is isomorphi to F. Letf : F

3

(x;y)! F be theho-

momorphism determined by f(x) = e

1

, f(y) = e

2

. Beause homomorphisms

behave well on assoiators, namely f((a;b;)) = (f(a);f(b);f()), the alula-

tion in the previousparagraphshowsthat f(u

1 ) =e

3 , f(u

2 ) =e

4 , f(v

1 ) =e

5 ,

f(v

2 )=e

6 ,f(v

3 )=e

7

andf(v

4 )=e

8

. ByLemma4.2,anyelementwofF

3 (x;y)

an be written as w = (x a1

y a2

u a3

u a4

)v a5

v a6

v a7

v a8

, and it now follows that

(15)

f(w)=(a

1

;a

2

;a

3

;a

4

;a

5

;a

6

;a

7

;a

8

). This meansthat f is ontoF, andalso that

theexponentsa

i

inthedeompositionofwareuniquelydeterminedbyw. Hene

f :F

3

(x;y)!F isanisomorphism.

WeonludethepaperwithsomestruturalinformationaboutF

3 (x;y).

Proposition5.5. LetQ=F

3

(x;y)beidentiedwith(Z 8

;)asinTheorem5.4.

Then A(Q) = N

(Q) = 00Z 6

and N

(Q) = N

(Q) = N(Q) = Z(Q) =

0000Z 4

.

Proof: We already know from Lemma 5.1 that A(Q) N

(Q). ByProposi-

tion2.8andLemmas3.2,4.1,

(x;x a1

y a2

;y)=(x;x a1

;y)(x;y a2

;y)

((x;x a

1

;y);x a

1

;y a

2

)((x;y a

2

;y);y a

2

;x a

1

)((x;x a

1

;y);y a

2

;x)

((x;y a2

;y);x a1

;x)((x;x a1

;y);y a2

;y)((x;y a2

;y);x a1

;y)

=(x;x a

1

;y)(x;y a

2

;y)v a

2

1 a

2 2a

1 a

2

2

v a

1 a

2

2 2a

1 a

2

3

=u a1

1 (u

1

;x;x) (a

1 )+(a

1 )

(u

1

;x;y) (a

1 )

u a2

2 (u

2

;y;y) (a

2 )+(a

2 )

(u

2

;y;x) (a

2 )

v a

2

1

a2 2a1a2

2

v a1a

2

2 2a1a2

3

=u a1

1 u

a2

2 v

(a1) (a1)

1

v (a

1 ) a

2

1 a

2 2a

1 a

2

2

v (a

2 ) a

1 a

2

2 2a

1 a

2

3

v

(a2) (a2)

4

:

Thus, ifeither a

1

6= 0or a

2

6=0then r=(x;x a

1

y a

2

;y)6=1. In otherwords,if

r2N

(Q)thenr200Z 6

. Weonlude,A(Q)N

(Q)00Z

6

A(Q),

soN

(Q)=A(Q)=00Z 6

.

SineQhasnilpotenylass3,wehave0000Z 4

Z(Q)N

(Q)=

N

(Q). Now,

(x;x;x a1

y a2

u a

3

1 u

a

4

2

)=(x;x;x a1

y a2

)(x;x;u a

3

1 u

a

4

2

)=(x;x;x a1

y a2

)v a

3

1 v

a

4

2

=[(x;x;y a2

)((x;x;y a2

);y a2

;x a1

)((x;x;y a2

);x a1

;x)((x;x;y a2

);x a1

;x)℄v a

3

1 v

a

4

2

=(x;x;y a

2

)v

a3 2a1a2

1

v a

4 a

1 a

2

2

2

=u a2

2 (u

2

;y;y) (a2)

(u

2

;y;x) 2(a2)

v

a3 2a1a2

1

v a4 a1a

2

2

2

=u a

2

2 v

a

3 2a

1 a

2

1

v a4 a1a

2

2

2 v

2(a2)

3 v

(a2)

4 :

If (x;x;x a1

y a2

u a

3

1 u

a

4

2

) =1,a

2

must be zero. Then v a

3

1 v

a

4

2

=1and thus a

3

=

a

4

=0. Therefore,if(x;x;x a

1

y a

2

w a

3

t a

4

)=1thena

2

=a

3

=a

4

=0. Finally,

(y;x;x a1

)=u a

1

1 (u

1

1

;x;x)

(a1)+(a1)

(u 1

1

;x;y) (a1)

=u a

1

1 v

(a

1 )+(a

1 )

1

v (a

1 )

2 :

So,(y;x;x a1

)=1implies a

1

=0. Summarizing,x a1

y a2

u a3

1 u

a4

2 2N

(Q) ifand

onlyifa

1

=a

2

=a

3

=a

4

=0. HeneN

(Q)=Z(Q)=0000Z 4

.

(16)

Referenes

[1℄ BarrosD.,Commutativeautomorphiloops,PhDdissertation,UniversityofSaoPaulo,in

preparation.

[2℄ BarrosD.,GrishkovA.,Vojtehovsk yP.,Commutativeautomorphi loopsoforderp 3

,J.

AlgebraAppl.,toappear.

[3℄ BrukR.H.,ASurveyofBinarySystems,Springer,1971.

[4℄ BrukR.H.,PaigeL.J.,Loops whoseinnermappingsareautomorphisms,Ann.ofMath.

(2)63(1956),308{323.

[5℄ Csorg}oP.,Themultipliationgroupofaniteommutativeautomorphiloopof orderof

powerofanoddprimepisap-group,J.Algebra350(2012),no.1,77{83.

[6℄ JedlikaP.,KinyonM.,Vojtehovsk yP.,Thestrutureofommutativeautomorphiloops,

Trans.Amer.Math.So.363(2011),no.1,365{384.

[7℄ JedlikaP.,KinyonM.,Vojtehovsk yP.,Construtionsofommutativeautomorphiloops,

Comm.Algebra38(2010),no.9,3243{3267.

[8℄ JedlikaP.,KinyonM.,Vojtehovsk yP.,Nilpotenyinautomorphi loopsofprimepower

order,J.Algebra350(2012),no.1,64{76.

[9℄ Johnson K.W., Kinyon M.K., Nagy G.P., Vojtehovsk y P., Searhing for small simple

automorphiloops,LMSJ.Comput.Math.14(2011),200{213.

[10℄ GrishkovA.N.,ShestakovI.P., Commutative Moufang loops and alternative algebras,J.

Algebra333(2011),1{13.

[11℄ Kinyon M.K., Kunen K., Phillips J.D., Vojtehovsk y P., The struture of automorphi

loops,inpreparation.

[12℄ WolframResearh,In., Mathematia,version8.0, WolframResearh,In., Champaign,

Illinois,2010.

D.Barros,A.Grishkov:

Instituteof Mathematis and Statistis, Universityof Sao Paulo, Ruado

Mat~

ao, 1010,Cidade

Universit

aria, S~

aoPaulo,SP,Brazil,CEP 05508-090

E-mail: dyleneime.usp.br

shuragrigmail.om

P.Vojtehovsk y:

DepartmentofMathematis,UniversityofDenver,2360SGaylordSt,Den-

ver,Colorado80208, USA

E-mail: petrmath.du.edu

(Reeived Marh31,2012 , revised April26,2012 )

参照

関連したドキュメント

Sommerville [10] classified the edge-to-edge monohedral tilings of the sphere with isosceles triangles, and those with scalene triangles in which the angles meeting at any one

That G requires infinitely nmny colors follows from Theorem 2 below, although it also follows directly from the fact that, for j > i, the i-th column contains a vertex adjacent

Dragomir, Trapezoidal-type Rules from an Inequali- ties Point of View,Handbook of Analytic-Computational Methods in Applied Mathematics, Editor: G.. Peˇ cari´c, On Euler

An iterative method for solving single variable nonlinear equation fx 0, with n 1, n ≥ 1, evaluations per iteration reaches to the maximum order of convergence 2 n and the

Several tiles can be related to a given digraph G (possibly with different diameters,) however the tile generated by the previous procedure has the same diameter as G... We will

In this paper we introduce the concept of free abstract fuzzy economy and use Kim and Lee’s existence theorems of best proximity pairs [14] and Wu’s existence theorem of

The main purpose of this paper is to show that all of the sequences generated by pairs of distinct elements in arbitrary finite algebra of a variety with the property (2, q) have

In this paper we study the regularity properties of solutions to a certain type of free boundary problems, resembling the obstacle problem but with no sign assumption, i.e., with