A Note On The Derived Length of The Unit Group

This article was downloaded by: [Indian Institute of Technology Roorkee]
On: 30 July 2011, At: 03:49

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Communications in Algebra
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/lagb20
A NOTE ON THE DERIVED LENGTH OF THE UNIT GROUP

OF A MODULAR GROUP ALGEBRA
a
Czesław Bagiński
a
Institute of Computer Science, Białystok Technical University, Wiejska 45A, Białystok,
15-351, Poland
Available online: 19 Aug 2006
To cite this article: Czesław Bagiński (2002): A NOTE ON THE DERIVED LENGTH OF THE UNIT GROUP OF A MODULAR GROUP
ALGEBRA, Communications in Algebra, 30:10, 4905-4913
To link to this article: http://dx.doi.org/10.1081/AGB-120014675
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions
This article may be used for research, teaching and private study purposes. Any substantial or systematic
reproduction, re-distribution, re-selling, loan, sub-licensing, systematic supply or distribution in any form to
anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2002 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
COMMUNICATIONS IN ALGEBRA
Vol. 30, No. 10, pp. 4905–4913, 2002
Downloaded by [Indian Institute of Technology Roorkee] at 03:49 30 July 2011
A NOTE ON THE DERIVED LENGTH

OF THE UNIT GROUP OF A MODULAR
GROUP ALGEBRA
Czesl⁄ aw Bagiński
Institute of Computer Science, Białystok Technical

University, Wiejska 45A, 15-351 Białystok, Poland
ABSTRACT
Let G be a finite p-group, p > 2; and let F be a field of char-

acteristic p: In this paper we show that if the commutator
subgroup of G is cyclic of order pk then the class of solvability
of the unit group of the group algebra FG is equal to
dlog2 ðpk þ 1Þe.
1. INTRODUCTION
Let p be a prime and let F be a field of characteristic p; p > 0: For a

finite group G let UðGÞ be the group of all units of the group algebra FG: If
UðGÞ is nilpotent there is a satisfactory description of its nilpotency class.[7]
Much less is known about the solvability class of UðGÞ (dl ðUðGÞÞ) when
4905
DOI: 10.1081/AGB-120014675 0092-7872 (Print); 1532-4125 (Online)

Copyright # 2002 by Marcel Dekker, Inc. www.dekker.com
4906 BAGIŃSKI
UðGÞ is solvable. First results concerning the derived length of the unit group
in this case were obtained by Shalev in.[6] He proved there the following:
Theorem 1. Let G be a finite group and let F be a field of characteristic p.
ðiÞ If p > 3; then UðGÞ is meta-abelian if and only if G is abelian.

ðiiÞ If p ¼ 3; then UðGÞ is meta-abelian if and only if G is either abelian
or nilpotent with jG0 j ¼ 3:

The description of finite groups for which UðGÞ is meta-abelian was com-
pleted for p ¼ 2 by Coleman and Sandling in[1] and independently by
Kurdics in[2]. The author does not know other results concerning description
of the derived length of UðGÞ:
Let us remind that, if p > 2 and UðGÞ is solvable, then G has a normal
Sylow p-subgroup P such that G=P is abelian or a 2-group.[3] So such groups
are very close to finite p-groups. In this paper we consider the solvability
class of UðGÞ when G is a finite p-group. The main aim of this paper is to
prove the following result.
Theorem 2. Let G be a finite p-group and let F be a field of characteristic p;

p > 2. If the commutator subgroup G0 of G is cyclic then the derived length of
UðGÞ is equal to dlog2 ðj G0 j þ1Þe:
As an easy consequence of the theorem we get the following strengthening
of Theorem 1 for p-groups of odd order.
Corollary 1. Let F be a field of characteristic p; p > 2; and let G be a finite

nonabelian p-group. Then the derived length of UðGÞ is not smaller than
dlog2 ðp þ 1Þe:
The notation is mostly standard. In particular, DðGÞ denotes the

augmentation ideal of the group algebra FG. If G is a finite p-group then
DðGÞ is nilpotent and we set tðGÞ ¼ minfn 2 N : DðGÞn ¼ 0g. By definition
DðGÞ0 ¼ FG: VðGÞ ¼ 1 þ DðGÞ stands for the group of normalized units of
FG; FðGÞ is the Frattini subgroup of G and gi G denotes the i-th term of its
lower central series, (g2 G ¼ G0 ), dl ðGÞ is the solvability class of G. We use
ðx; yÞ ¼ x1 y1 xy for the group commutator of elements x and y of a group.
The group commutator of weight n; n > 2; we define inductively by
ðx1 ; x2 ; . . . ; xn Þ ¼ ððx1 ; x2 ; . . . ; xn1 Þ; xn Þ: For elements a and b of the group
algebra FG we use ½a; b ¼ ab ba to denote the Lie commutators and
inductively ½a1 ; a2 ; . . . ; an ¼ ½½a1 ; a2 ; . . . ; an1 ; an for n > 2: For subsets A
and B of FG we write ½A; B for the subspace of FG spanned by all ½a; b;
where a 2 A and b 2 B: If r is a real number then by dre we denote the
minimal integer not smaller than r:
UNIT GROUP OF MODULAR GROUP ALGEBRA 4907
2. PROOF OF THE MAIN RESULT
For the sake of clearity we shall prove Theorem 2 in a sequence of

lemmas. In what follows we assume that F is a field of characteristic p and G
is a finite p-group. We begin with an observation of independent interest
describing linear bases of certain spaces related to the ideal FGDðHÞ, where
H is a fixed normal subgroup of G.
Lemma 1. Let H be a normal subgroup of a finite p-group G and let a ! a be

the natural epimorphism of FG onto FðG=HÞ: Let am1 ; . . . ; amsm 2 FG be such
that am1 ; . . . ;
amsm is a linear basis of DðG=HÞm modulo DðG=HÞmþ1 and let
bn1 ; . . . ; bntn be a basis of DðHÞn modulo DðHÞnþ1 : Then the elements ami bnj ;
i ¼ 1; . . . ; sm ; j ¼ 1; . . . ; tn ; form a basis of DðGÞm DðHÞn þ FGDðHÞnþ1 modulo
DðGÞmþ1 DðHÞn þ FGDðHÞnþ1 :
Proof. Let V be a linear subspace of FG such that dim V ¼ jG=Hj and
V ¼ FðG=HÞ: Since FGDðHÞ is the kernel of the natural epimorphism of FG
onto FðG=HÞ, we have V \ FGDðHÞ ¼ 0 and then FG ¼ V FGDðHÞ:
Hence
FG ¼ V ðV FGDðHÞÞDðHÞ ¼ V ðV ðV FGDðHÞÞÞDðHÞ
¼ V ðVDðHÞ þ VDðHÞ2 þ VDðHÞ3 . . .Þ
¼ V VDðHÞ:
Multiplying FG by DðHÞn on the right hand side we obtain then
FGDðHÞn ¼ VDðHÞn :
Let aki ; 0 k < tðG=HÞ; 1 i sk form a basis of V such that the subspace
Vm ¼ V \ DðGÞm is spanned by all aki with k % m: Then obviously DðGÞm þ
FGDðHÞ ¼ Vm VDðHÞ: Therefore
DðGÞm DðHÞn þ FGDðHÞnþ1 ¼ Vm DðHÞn þ VDðHÞnþ1 : ð1Þ
Let blj ; 1 l < tðHÞ; 1 j tl form a basis (over F) of DðHÞ such that the
subspace DðHÞn is spanned by all blj with l % n: It is clear that all the pro-
ducts ami dnj span FGDðHÞ: The number of these products equals jG=Hj
dim DðHÞ ¼ jG=HjðjHj 1Þ ¼ dim FGDðHÞ: So they form a basis of
FGDðHÞ: Now, it is easily seen by (1) that for fixed m and n the set A ¼
faki blj : ðk % m and l % nÞ or ðk % 0 and l % n þ 1Þg is a basis of DðGÞm
4908 BAGIŃSKI
DðHÞn þ FGDðHÞnþ1 and its subset B ¼ faki blj : ðk % m þ 1 and l % nÞ

or ðk % 0 and l % n þ 1Þg is a basis of DðGÞmþ1 DðHÞn þ FGDðHÞnþ1 :
Hence AnB ¼ fami bnj : 1 i sm ; 1 j tn g is a basis of DðGÞm DðHÞn þ
FGDðHÞnþ1 modulo DðGÞmþ1 DðHÞn þ FGDðHÞnþ1 and the lemma follows.
The next lemma can be easily derived from well known results. u
Lemma 2. Let H be a normal subgroup of a finite p-group G: Then

DðGÞDðHÞ ¼ DðHÞDðGÞ
if and only if ðG; HÞ FðHÞ:

Proof. By Lemma 2.2 of[4] we have G \ ð1 þ DðGÞDðHÞ þ DðHÞDðGÞÞ ¼
ðG; HÞHp : Moreover it follows easily from the proof of Prop. III.1.15(ii) of[5]
that G \ ð1 þ DðGÞDðHÞÞ ¼ FðHÞ: Hence DðGÞDðHÞ ¼ DðHÞDðGÞ if and
only if ðG; HÞHp ¼ FðHÞ: But Hp FðHÞ; so the lemma is proved. u
Lemma 3. Let G be a finite p-group such that for i % 1
giþ1 G ðgi GÞp : ð2Þ
Then for all j; k; m; n % 1

m k k1 0 mþ1
ðiÞ D G0 ; D G D G D G ;
j k jþk2 0
ðiiÞ D G ; D G D G D G ;
j 0 m k 0 n jþk2 0 mþnþ1
ðiiiÞ D G D G ; D G D G D G D G :
Proof. The proof depends on rather standard calculations and goes by

induction used several times. Note first that by (2) we have
2
½a; b; g 2 FGDðG0 Þ ð3Þ
for all a; b; g 2 FG: In fact, if x; y; z 2 G then
½x; y ¼ xy yx ¼ yxððx; yÞ 1Þ ð4Þ
and
½x; y; z ¼ ½yxððx; yÞ 1Þ; z ¼ yx½ðx; yÞ; z þ ½yx; zððx; yÞ 1Þ

¼ yxzðx; yÞððx; y; zÞ 1Þ þ zyxðð yx; zÞ 1Þððx; yÞ 1Þ
2 2
2 FGDðg3 GÞ þ FGDðG0 Þ FGDðG0 Þ ð5Þ
as Dðg3 GÞ DðG0p Þ DðG0 Þp DðG0 Þ2 : So we get (3) by linearity of Lie

brackets. Hence ½DðGÞ; DðGÞ; DðGÞ FGDðG0 Þ2 :
(i) It is easily seen that
2
½DðG0 Þ; DðGÞ FGDðG0 Þ : ð6Þ
Actually, for z ¼ ðx; yÞ 2 G0 and g 2 G we have


½z; g ¼ ½z 1; g ¼ x1 y1 ½½x; y; g þ x1 y1 ; g ½x; y
and for z ¼ ðx1 ; y1 Þðx2 ; y2 Þ ðxs ; ys Þ
½z; g ¼ ðx1 ; y1 Þ ðxs1 ; ys1 Þ½ðxs ; ys Þ; g þ h½ðx1 ; y1 Þ ðxs1 ; ys1 Þ; gh ;
where h ¼ ðxs ; ys Þ: Hence by (3) we get ½z; g 2 FGDðG0 Þ2 and by induction

on m
½DðG0 Þm ; DðGÞ DðG0 Þm1 ½DðG0 Þ; DðGÞ þ ½DðG0 Þm1 ; DðGÞDðG0 Þ

FGDðG0 Þmþ1 :
Therefore by induction on k we obtain
½DðG0 Þm ; DðGÞk DðGÞk1 ½DðG0 Þm ; DðGÞ þ ½DðG0 Þm ; DðGÞk1 DðGÞ

DðGÞk1 DðG0 Þmþ1 :
(ii) The proof is very similar to that of (i). For k ¼ 1 we have
½DðGÞ; DðGÞ FGDðG0 Þ DðGÞ0 DðG0 Þ
and by induction
½DðGÞj ; DðGÞ DðGÞj1 ½DðGÞ; DðGÞ þ ½DðGÞj1 ; DðGÞDðGÞ

DðGÞj1 DðG0 Þ:
Hence
½DðGÞj ; DðGÞk

DðGÞk1 DðGÞj ; DðGÞ þ DðGÞj ; DðGÞk1 DðGÞ
DðGÞjþk2 DðG0 Þ þ DðGÞjþk3 DðG0 ÞDðGÞ DðGÞjþk2 DðG0 Þ:
4910 BAGIŃSKI
2
(iii) By (2) we have ðG0 ; G0 Þ g4 G ðG0 Þp so by (ii) we obtain
m n mþn2 mþn2 p2
DðG0 Þ ; DðG0 Þ DðG0 Þ Dðg4 GÞ DðG0 Þ DðG0 Þ
mþnþ1
DðG0 Þ :
Hence
m n
DðGÞj DðG0 Þ ; DðGÞk DðG0 Þ
m n m n
DðGÞjþk DðG0 Þ ; DðG0 Þ þ DðGÞj DðG0 Þ ; DðGÞk DðG0 Þ
n m mþn
þ DðGÞk DðGÞj ; DðG0 Þ DðG0 Þ þ DðGÞj ; DðGÞk DðG0 Þ :
Now by (i) and (ii) it is seen that all the summands appearing on the right
hand side are contained in the ideal DðGÞjþk2 DðG0 Þmþnþ1 : This ends the
proof of the lemma. u
Let d0 ðFGÞ ¼ FG and for n % 1 let dn ðFGÞ be the two sided (associa-
tive) ideal of FG generated by all Lie commutators ½a; b; where a; b 2
dn1 ðFGÞ: By dl ðFGÞ we denote the minimal integer n such that dn ðFGÞ ¼ 0:
As an immediate consequence of Lemma 3 we obtain the following
corollaries which strengthen Lemma 2.1 and Corollary 2.2 of [6].
Corollary 4. If G is an finite p-group such that for all i % 2 giþ1 G ðgi GÞp ; then
dn ðFGÞ " FGDðG0 Þ2 1 : In particular dlðFGÞ dlog2 ðtðG0 Þ þ 1Þe:
Proof. It is obvious that FGDðG0 Þm ¼ DðGÞDðG0 Þm þDðG0 Þm : Hence by

Lemma 3
½FGDðG0 Þk ; FGDðG0 Þm DðG0 Þkþmþ1

n n nþ1
and then ½FGDðG0 Þ2 1 ; FGDðG0 Þ2 1 DðG0 Þ2 1 . Now, let n be an ninteger
equal to dlog2 ðtðG0 Þ þ 1Þe: We have n % log2 ðtðG0 Þ þ 1Þ and DðG0 Þ2 1 ¼ 0
that is dlðFGÞ n: u
Corollary 5. If G is a finite p-group such that for all i % 2; giþ1 G ðgi GÞp ;
then the n-th term
n
of the derived series of UðGÞ is contained in the subgroup
1 þ FGDðG0 Þ2 1 : In particular the derived length of UðGÞ is not bigger than
dlog2 ðtðG0 Þ þ 1Þe:
Proof. Let a; b 2 UðGÞ: Then by the equality ða; bÞ ¼ 1 þ a1 b1 ½a; b the
m-th term of the derived series of UðGÞ is contained in 1 þ dm ðFGÞ: Hence
the result follows from Corollary 4. u
Proposition 6. Let G be a finite p-group, p % 2; with a cyclic commutator

subgroup G0 ¼ hui: Then dl ð1 þ DðGÞ2 Þ % dlog2 tðG0 Þ þ 1e: In particular
dl UðGÞ % dlog2 tðG0 Þ þ 1e:
Proof. Let x; y 2 G be such elements that ðx; yÞ ¼ u: It suffices to prove the

assertion for the case G ¼ hx; yi: Let
a0 ¼ 1 þ ðx 1Þ2 ; b0 ¼ 1 þ ð y 1Þ2 ; g0 ¼ 1 þ ðx 1Þð y 1Þ
and by induction
an ¼ ðan1 ; gn1 Þ; bn ¼ ðbn1 ; gn1 Þ; gn ¼ ðan1 ; bn1 Þ:
Since ½x; y ¼ ðu 1Þ þ ð yx 1Þðu 1Þ and ½x; y; z 2 FGðu 1Þ2 ; for every

z 2 G; by (4) and (5), we have
½a0 ; b0 ¼ ½ðx 1Þ2 ; ð y 1Þ2 ¼ ðx 1Þð y 1Þ½x; y

þ ðx 1Þ½x; yð y 1Þ þ ð y 1Þ½x; yðx 1Þ þ ½x; yð y 1Þðx 1Þ
Z 4ðx 1Þð y 1Þðu 1Þðmod DðGÞ3 ðu 1Þ þ FGðu 1Þ2 Þ;
½a0 ; g0 ¼ ½ðx 1Þ2 ; ðx 1Þð y 1Þ ¼ ðx 1Þ2 ½x; y

þ ðx 1Þ½x; yðx 1Þ ¼ 2ðx 1Þ2 ½x; y þ ½x; y; z
Z 2ðx 1Þ2 ðu 1Þðmod DðGÞ3 ðu 1Þ þ FGDðu 1Þ2 Þ:
and similarly
½b0 ; g0 Z 2ð y 1Þ2 ðu 1Þðmod DðGÞ3 ðu 1Þ þ FGDðu 1Þ2 Þ:
Hence
1

g1 ¼ ða0 ; b0 Þ ¼ 1 þ ½a0 ; b0 þ a1
0 b0 1 ½a0 ; b0
Z 1 þ ½a0 ; b0 ðmod DðGÞ3 ðu 1Þ þ FGðu 1Þ2 Þ

Z 1 þ 4ðg0 1Þðu 1Þðmod DðGÞ3 ðu 1Þ þ FGðu 1Þ2 Þ
and
a1 Z 1 þ 2ða0 1Þðu 1Þðmod DðGÞ3 ðu 1Þ þ FGðu 1Þ2 Þ;

b1 Z 1 þ 2ðb0 1Þðu 1Þðmod DðGÞ3 ðu 1Þ þ FGðu 1Þ2 Þ:
4912 BAGIŃSKI
Now assume for the induction step that

n
an Z 1 þ an ða0 1Þðu 1Þ2 1 þa0n ;
n
1
bn Z 1 þ bn ðb0 1Þðu 1Þ2 þb0n ;
n
gn Z 1 þ cn ðg0 1Þðu 1Þ2 1 þg0n ;
n n
where a0n ; b0n ; g0n 2 In ¼ DðGÞ3 ðu 1Þ2 1 þ FGðu 1Þ2 and an ; bn ; cn are
powers of 2 with an ¼ bn : Obviously an ; bn ; cn are not 0 in F nas char F ¼ p > 2:
Moreover by Lemma 1, an ; bn ; gn Z= 1ðmod In Þ if ðu 1Þ2 1 6¼ 0: It follows
from Lemma (3) that for a 2 fan ; bn ; gn ; a0n ; b0n ; g0n g and b 2 fa0n ; b0n ; g0n g, ½a; b 2
Inþ1 : Therefore
n n
1
½an ; gn ¼ an cn ½ða0 1Þðu 1Þ2 ; ðg0 1Þðu 1Þ2 1
nþ1
2
Z an cn ½ðx 1Þ2 ; ðx 1Þð y 1Þðu 1Þ2
nþ1
1
Z 2an cn ðx 1Þ2 ðu 1Þ2
nþ1
1
Z 2an cn ða0 1Þðu 1Þ2 ðmod Inþ1 Þ
nþ1
and similarly ½bn ; gn Z 2bn cn ðbn 1Þðu 1Þ2 1 ðmod Inþ1 Þ and ½an ; bn Z
nþ1
ðg0 1Þðu 1Þ2 1 ðmod Inþ1 Þ: Thus similarly as in the case n ¼ 0
nþ1
1
anþ1 Z 1 þ 2ða0 1Þðu 1Þ2 ðmod Inþ1 Þ;
2nþ1 1
bnþ1 Z 1 þ 2ðb0 1Þðu 1Þ ðmod Inþ1 Þ;
2nþ1 1
gnþ1 Z 1 þ 2ðg0 1Þðu 1Þ ðmod Inþ1 Þ:
This ends the proof of the induction step. Therefore for 2n 1 < tðG0 Þ the n-
th term of the derived series of 1 þ DðGÞ2 is not trivial. This ends the proof
of the proposition and Theorem 2. u
REFERENCES
1. Coleman, D.B.; Sandling, R. Mod 2 Group Algebras with Metabelian

Unit Groups. Preprints of the University of Manchester 1994, 94 (6).
2. Kurdics, J. On Group Algebras with Metabelian Unit Groups. Periodica
Mathematica Hungarica 1996, 32, 57–64.
3. Passi, I.B.S.; Passman, D.S.; Sehgal, S.K. Lie Solvable Group Rings.
Canad. J. Math. 1973, 25, 748–757.
4. Sandling, R. The Modular Group Algebra of a Central-elementary-by-
abelian p-group. Arch. Math. 1989, 52, 22–27.
5. Sehgal, S.K. Topics in Group Rings; Marcel Dekker: New York, 1978.
6. Shalev, A. Meta-abelian Unit Groups of Group Algebras are Usually
Abelian. J. Pure and Appl. Algebra 1991, 72, 295–302.
7. Shalev, A. The Nilpotency Class of the Unit Group of a Modular Group

Algebra III. Arch. Math. 1993, 60, 136–145.
Received May 2001

Revised November 2001

A Note On The Derived Length of The Unit Group

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Note On The Derived Length of The Unit Group

Uploaded by

Copyright:

Available Formats

This article was downloaded by: [Indian Institute of Technology Roorkee]

On: 30 July 2011, At: 03:49

A NOTE ON THE DERIVED LENGTH OF THE UNIT GROUP

Available online: 19 Aug 2006

To link to this article: http://dx.doi.org/10.1081/AGB-120014675

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

A NOTE ON THE DERIVED LENGTH

Institute of Computer Science, Białystok Technical

Let G be a ﬁnite p-group, p > 2; and let F be a ﬁeld of char-

Let p be a prime and let F be a ﬁeld of characteristic p; p > 0: For a

DOI: 10.1081/AGB-120014675 0092-7872 (Print); 1532-4125 (Online)

ðiÞ If p > 3; then UðGÞ is meta-abelian if and only if G is abelian.

or nilpotent with jG0 j ¼ 3:

Theorem 2. Let G be a ﬁnite p-group and let F be a ﬁeld of characteristic p;

Corollary 1. Let F be a ﬁeld of characteristic p; p > 2; and let G be a ﬁnite

The notation is mostly standard. In particular, DðGÞ denotes the

2. PROOF OF THE MAIN RESULT

For the sake of clearity we shall prove Theorem 2 in a sequence of

Lemma 1. Let H be a normal subgroup of a ﬁnite p-group G and let a ! a be

Multiplying FG by DðHÞn on the right hand side we obtain then

DðGÞm DðHÞn þ FGDðHÞnþ1 ¼ Vm DðHÞn þ VDðHÞnþ1 : ð1Þ

DðHÞn þ FGDðHÞnþ1 and its subset B ¼ faki blj : ðk % m þ 1 and l % nÞ

Lemma 2. Let H be a normal subgroup of a ﬁnite p-group G: Then

if and only if ðG; HÞ FðHÞ:

Lemma 3. Let G be a ﬁnite p-group such that for i % 1

giþ1 G ðgi GÞp : ð2Þ

Then for all j; k; m; n % 1

Proof. The proof depends on rather standard calculations and goes by

for all a; b; g 2 FG: In fact, if x; y; z 2 G then

½x; y ¼ xy yx ¼ yxððx; yÞ 1Þ ð4Þ

½x; y; z ¼ ½yxððx; yÞ 1Þ; z ¼ yx½ðx; yÞ; z þ ½yx; zððx; yÞ 1Þ

as Dðg3 GÞ  DðG0p Þ  DðG0 Þp  DðG0 Þ2 : So we get (3) by linearity of Lie

Actually, for z ¼ ðx; yÞ 2 G0 and g 2 G we have

and for z ¼ ðx1 ; y1 Þðx2 ; y2 Þ    ðxs ; ys Þ

½z; g ¼ ðx1 ; y1 Þ    ðxs1 ; ys1 Þ½ðxs ; ys Þ; g þ h½ðx1 ; y1 Þ    ðxs1 ; ys1 Þ; gh ;

where h ¼ ðxs ; ys Þ: Hence by (3) we get ½z; g 2 FGDðG0 Þ2 and by induction

½DðG0 Þm ; DðGÞ  DðG0 Þm1 ½DðG0 Þ; DðGÞ þ ½DðG0 Þm1 ; DðGÞDðG0 Þ 

Therefore by induction on k we obtain

½DðG0 Þm ; DðGÞk   DðGÞk1 ½DðG0 Þm ; DðGÞ þ ½DðG0 Þm ; DðGÞk1 DðGÞ

(ii) The proof is very similar to that of (i). For k ¼ 1 we have

½DðGÞ; DðGÞ  FGDðG0 Þ  DðGÞ0 DðG0 Þ

½DðGÞj ; DðGÞ  DðGÞj1 ½DðGÞ; DðGÞ þ ½DðGÞj1 ; DðGÞDðGÞ

Proof. It is obvious that FGDðG0 Þm ¼ DðGÞDðG0 Þm þDðG0 Þm : Hence by

½FGDðG0 Þk ; FGDðG0 Þm   DðG0 Þkþmþ1

Proposition 6. Let G be a ﬁnite p-group, p % 2; with a cyclic commutator

Proof. Let x; y 2 G be such elements that ðx; yÞ ¼ u: It suﬃces to prove the

a0 ¼ 1 þ ðx 1Þ2 ; b0 ¼ 1 þ ð y 1Þ2 ; g0 ¼ 1 þ ðx 1Þð y 1Þ

an ¼ ðan1 ; gn1 Þ; bn ¼ ðbn1 ; gn1 Þ; gn ¼ ðan1 ; bn1 Þ:

Since ½x; y ¼ ðu 1Þ þ ð yx 1Þðu 1Þ and ½x; y; z 2 FGðu 1Þ2 ; for every

½a0 ; b0  ¼ ½ðx 1Þ2 ; ð y 1Þ2  ¼ ðx 1Þð y 1Þ½x; y

½a0 ; g0  ¼ ½ðx 1Þ2 ; ðx 1Þð y 1Þ ¼ ðx 1Þ2 ½x; y

½b0 ; g0  Z 2ð y 1Þ2 ðu 1Þðmod DðGÞ3 ðu 1Þ þ FGDðu 1Þ2 Þ:

Z 1 þ ½a0 ; b0 ðmod DðGÞ3 ðu 1Þ þ FGðu 1Þ2 Þ

a1 Z 1 þ 2ða0 1Þðu 1Þðmod DðGÞ3 ðu 1Þ þ FGðu 1Þ2 Þ;

Now assume for the induction step that

1. Coleman, D.B.; Sandling, R. Mod 2 Group Algebras with Metabelian

7. Shalev, A. The Nilpotency Class of the Unit Group of a Modular Group

Received May 2001

You might also like

½x; y ¼ xy yx ¼ yxððx; yÞ 1Þ ð4Þ

½x; y; z ¼ ½yxððx; yÞ 1Þ; z ¼ yx½ðx; yÞ; z þ ½yx; zððx; yÞ 1Þ

as Dðg3 GÞ DðG0p Þ DðG0 Þp DðG0 Þ2 : So we get (3) by linearity of Lie

and for z ¼ ðx1 ; y1 Þðx2 ; y2 Þ ðxs ; ys Þ

½z; g ¼ ðx1 ; y1 Þ ðxs1 ; ys1 Þ½ðxs ; ys Þ; g þ h½ðx1 ; y1 Þ ðxs1 ; ys1 Þ; gh ;

where h ¼ ðxs ; ys Þ: Hence by (3) we get ½z; g 2 FGDðG0 Þ2 and by induction

½DðG0 Þm ; DðGÞ DðG0 Þm1 ½DðG0 Þ; DðGÞ þ ½DðG0 Þm1 ; DðGÞDðG0 Þ

½DðG0 Þm ; DðGÞk DðGÞk1 ½DðG0 Þm ; DðGÞ þ ½DðG0 Þm ; DðGÞk1 DðGÞ

½DðGÞ; DðGÞ FGDðG0 Þ DðGÞ0 DðG0 Þ

½DðGÞj ; DðGÞ DðGÞj1 ½DðGÞ; DðGÞ þ ½DðGÞj1 ; DðGÞDðGÞ

½FGDðG0 Þk ; FGDðG0 Þm DðG0 Þkþmþ1

Since ½x; y ¼ ðu 1Þ þ ð yx 1Þðu 1Þ and ½x; y; z 2 FGðu 1Þ2 ; for every

½a0 ; b0 ¼ ½ðx 1Þ2 ; ð y 1Þ2 ¼ ðx 1Þð y 1Þ½x; y

½a0 ; g0 ¼ ½ðx 1Þ2 ; ðx 1Þð y 1Þ ¼ ðx 1Þ2 ½x; y

½b0 ; g0 Z 2ð y 1Þ2 ðu 1Þðmod DðGÞ3 ðu 1Þ þ FGDðu 1Þ2 Þ:

Z 1 þ ½a0 ; b0 ðmod DðGÞ3 ðu 1Þ þ FGðu 1Þ2 Þ