theorems like sylow's (hall)

Upload: michel57

Post on 07-Jul-2018

219 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/18/2019 Theorems Like Sylow's (Hall)

    1/19

    THEOREMS LIKE SYLOW'S

    By

      P.

      HALL

    [Received  24  June 1955.—Read 17 November 1955]

    1. Discussion

     of

     results

    1.1.  Notation.

     Let

     w

     be any set of

     primes

     and let

     w'

     be the

     comple-

    mentary

      set

      consisting

     of all

     primes

     not in

     w.  Ev ery positive integer

     m

    can be  expressed uniquely in the  form

    m

      =

      m

    m

    m

    VT

    ,,

    where m

    m

     is the  largest divisor of m which has no prime divisor in m'.

    We denote the order  of a  finite group  G by  (G).  If  (6%, =  (G ) we call

    G a  w-group.  If

     H

     is a subgroup of G such tha t

      (H) = (G)^,

     we call

     H

     an

    S

    m

    -subgroup of  G.  In

      particular, when w  consists

     of a

      single prime p,

     an

    $p-subgroup

      is the

      same

      as a

      Sylow ^-subgroup.

      But for

      general

      -nr, a

    Sylow m-subgroup

     is

     not the same as an ^ - sub grou p , bu t

     is

     an $p-subgroup

    for some prime ^in-nr.

    Let E

    m)

      C

    m

    , and

     D

    m

     be the

      following propositions about

     a

     finite group G.

    E

    m

    : G has at  least one  ^ - s u b g r o u p .

    C

    m

    : G satisfies  E

    m

     and any two /S^-subgroups of G are conjugate  in G.

    D

    m

    :  G satisfies  C

    m

     and  every -nr-subgroup  of  G is  contained  in  some

    #

    OT

    -subgroup

      of

      G.

    Obviously  D  ̂can also be expressed  in either of the  alternative forms:

    (i)  G satisfies  C

    m

     and

      every maximal -nr-subgroup

     of  G is an

      #

    OT

    -subgroup

    of  G;

     (ii)

     G satisfies  E

    m

     and, if

     H

      is an

      ^ - s u b g r o u p

      of

      G

     and

     L

     is any

    xir-subgroup of  G, then

      L

      ^

      H

    x

      for  some

     x e G.

    1.2.  D-theorems. A  sufficient condition for a  finite group to  satisfy

      D

    w

    m ay  be  called  a  D-theorem. Similarly  for  ^-theorems  and  C-theorems.

    Our main object is to prove a new D-theorem,  viz. Theorem  D5, below.

    The two basic D-theorems are, of course, those due to Sylow and Schur,

    viz.

    THEOREM

     Dl.  If p  is a  prime, then every finite group sa tisfies D

    p

    .

    THEOREM

      D2. / /  G has a n ormal Abelian 8

    m

    .-subgroup, then G satisfies D

    m

    .

    We mak e once

     and for all the

     obvious remark th at,

     if

     m

    x

     and

     m

    2

     are two

    sets

     of

      primes such that  (G)

    mi

     — (G)^,  then

      for

      this particular group G

    the propositions  D

    mi

     and  J)

    m%

      are  equivalent:  G satisfies bo th or  neither.

    Similarly for the corresponding E-

     and

     C-propositions. So Sylow

    3

    s theorem

    Proc.  London Math.  Soc.  3) 6  1956)

  • 8/18/2019 Theorems Like Sylow's (Hall)

    2/19

    THE OR EM S LI K E SYLO W'S 287

    may be expressed by saying that  G  satisfies  D

    m

      whenever at most one

    prime in

      m

      divides

      (G).

    All -nr-groups satisfy  D

    m

      trivially.

    In an earlier paper (9), we proved

    THEOREM

      D 3.

      If G is soluble, then G satisfies D

    m

    for all

     m.

    Quite recently, W ielandt (14) has derived a remark able new D -theorem,

    which resembles Sylow's theorem and differs from D2 and D3 in the fact

    that it does not reduce to a triviality for simple groups.

    We define the propositions  E%,

      E^, 0%,

     and

      D

    s

    m

     abo ut a finite grou p

     G

    as follows.

    E^\ G has a ni lpotent ^- su bg ro up .

    E% f\ G  has a soluble ^- su bg ro up .

    C

    s

    m

    : G satisfies  C

    m

      and its /S^-subgroups are soluble.

    D ^ :  G satisfies  D^  and its -nr-subgroups are soluble.

    Wielandt's theorem is

    THEOREM   D4.  E%  implies D

    m

    .

    This makes it superfluous to introduce the propositions   C^. and  D

    7

    ^,

    analogous to

      C

    s

    m

      and

      D%

      but with 'soluble' replaced by 'nilpotent', since

    they would both be equivalent to

      E^.

    1.3. That  E% . does not in general imply  D

    m

     or even  C

    m

     is shown, as h as

    been remarked before, by the simple group of order 168 which has two

    distinct classes of conjugate o ctahed ral subgroups. This grou p satisfies

    Uf.3 but not O

    2

    ,3-

    Wielandt suggests as a possibility that the presence of a  supersoluble

    ^-subgroup might be sufficient to imply

      C

    m

      or perhaps even

      D

    m

    .

      B u t

    this is not the case, as is shown by the following examples.

    We denote by

     L

    p

      the simple group of order

     %p(p

    2

    —l),

     where

     p

     is a prim e

    greater tha n   3.  Then L

    lly

      of order 660, has two distinct classes of con jugate

    $

    2 3

    -subgroups, one class being dihedral (and therefore supersoluble) and

    the other class tetrah edra l. Again, L

    ei

      has two distinct classes of conjugate

    $

    2  3  5

    -subgroups, one class being dihedral and the other icosahedral (and

    therefore not even soluble). For these properties of  L

    p

    ,  see Burnside (1),

    Chapter XX.

    However, it is easy to see that  any two supersoluble  S

    m

    -subgroups must

    be conjugate.  A more general result m ay be described as follows. Le t

    p

    x

    , p

    2

    ,...,  p

    r

      be distinct primes. We say th at a finite group H  has a  Sylow

    series of complexion (p

    lt

      p

    2

    ,---,  p

    r

    )  if  (H )  is divisible by no primes other

    than  p

    1}

      p

    2

    , -,  p

    r

      and if, for each  i  = 1,

      2,...,

      r—1 H  has a normal (and

    therefore characteristic)

      8

    PltPt

      ^-subgroup. Then we may state :

  • 8/18/2019 Theorems Like Sylow's (Hall)

    3/19

    288 P. H A L L

    T H E O R E M

      A l .

      Let p

    v

      p^,...,  P

    r

      be the distinct primes which divide (G)

    m

    ,

    arranged in any given order. Then any two S

    m

    -subgroups of G both of w hich

    have a Sylow-series of complexion (p

    v

      p

    2

    ,--,  p

    r

    ) must be conjugate in 0.

    The conjugacy of supersoluble /S^-subgroups follows from this theorem,

    because a supersoluble group of order pi

    1

    p%

    i

    ...p%

    r

    , where p

    x

      >  p

    2

      > ... >  p

    r

    ,

    always has a Sylow series of complexion  (Pi,P2,---,p

    r

    )'

    The group

      L

    n

      does no t con tradict Theorem

      A1

      because its dihedral

    #

    2

     g-subgroups h ave Sylow series of complexion (3,2) bu t none of com-

    plexion (2,3), whereas its tetrahedral $

    2)3

    -subgroups have Sylow series of

    complexion (2,3) but none of complexion (3,2).

    I know of no counter-example to the conjecture that, if 2 does not

    belong to  m,  then  E

    m

      implies  D

    m

    .

    1.4. The principal theorem of the present pape r is

    THEOREM  D5 . / /

      K is a normal subgroup of 0 such that K satisfies Ei^

    and G/K satisfies D^, then G satisfies D% .

    Here we merely note a few corollaries of this theorem.

    A chain of subgroups

      G = G

    o

     >

      G

    t

      > ... >

      G

    r

     =

      1 of a group

      G

     will

    be called a series of  G  if each term  G

    i

     is norm al in the p receding te rm  G^

    v

    thoug h n ot necessarily norm al in

      G.

    C OR OL L AR Y

      D 5 . 1 . / /

      G has a series G = G

    Q

     >

      G

    x

      > ... >

      G

    r

      =

      1

      such

    that each of the factor groups

      G

    i

    _

    x

    jG

    i

      satisfies E

    1

    ^, then G satisfies D^.

    It would be sufficient here to postulate that the first factor group of the

    series,

      G/G

    1}

      satisfies the weaker condition

      D%

      instead of

     E%.

    Following Cunihin (2), we say that a finite group is  vr-separable  if all

    its c.f. (= composition factors) are Ttr'p-groups for various primes

     p

      in

      m.

    Thus  G  is tn--separable if and only if no two distinct primes in  m  divide

    the order of any c.f. of  G.  If p  is a prime, all finite grqmps are ^-separable

    If  G  is both -nr-separable and xo-'-separable, then  G is soluble: for th en the

    order of any c.f. of  G  must have the form  p

    a

    qP  with p  in TO- and  qin m'\

    but by a classical theorem of Burnside, groups of order  p^qP  are always

    soluble, so th a t either a = 1, fi = 0 or else a = 0, £ = 1.

    Clearly, if  m

    x

      is any subset of  m,  then every m-separable group is also

    •ro-j-separable. In (2), Theorem X I, Cunihin proves th a t all xtr-separable

    groups satisfy  C^,  and therefore, by the preceding remark, also  C

    s

    mx

      for

    any subset  TD-J of  m.  This result is contained in

    C OR OL L AR Y  D 5 . 2 .  All -m-separable groups satisfy

      D

    s

    mi

    ,

      for any subset

      TO^

    o

    m

  • 8/18/2019 Theorems Like Sylow's (Hall)

    4/19

    THEO REM S LIK E SYLOW 'S 289

    This corollary also contains as a special case the main Z)-theorem of

    Cunihin's paper, Theorem XII, (1) and (2).

    We call a finite group

      G m-serial

      if every c.f. of

      G

      is either a -nr-group

    or else a -ur'-group. Equivalently,  G  is -nr-serial if it has a  m-series,  i.e. a

    series  G =  G

    o

     >  G

    x

     > ... >  G

    r

     =  1 such tha t each of the factor groups

    G

    i

    _

    1

    IG

    i

     is eithe r a xtr-group or else a -nr'-group. Clearly, -nr-serial groups

    are the same as m'-serial groups. If p  is any prime in w and  q any prime

    in xtr', then every xtr-serial group is p,  ^-separable. Hence we have

    COROLLARY  D5.3.  All m-serial groups satisfy D ,

    pq

      for any p in m and

    any a in m'.

    We note that, by the theorem of Burnside already mentioned,  D

    p

     q

      is

    equivalent to  D

    s

    vq

    .

    1.5. In his well-known textbook (15), Chapter IV, Satz 27, Zassenhaus

    proves two C-theorems. The corresponding D-theorem s, which follow a t

    once,

      are

    THEOREM

     D6. / /  G has a normal

     8^-subgroup K such that G/K is soluble,

    then G satisfies D

    m

    .

    THEOREM   D7.  If G has a normal

     S

    m

    >-subgroup

      K such that K is soluble,

    then G satisfies D

    m

    .

    We call a finite group  G vx-soluble  if it is both -nr-separable and xo--serial.

    Equivalently,  G  is xtr-soluble if every c.f. of  G  is either a ^p-group (and

    therefore cyclic of order p)  for some p  in  vx,  or else a xcr'-group.t

    In (3), Cunihin proves the following generalizations of D6 and D7.

    THEOREM  D6.*  Every m-soluble group satisfies D^.

    1

     for any subset m

    x

     o f -m .

    THEOREM

     D7.*

      Every m-soluble group satisfies D^Jor any subset TJT

    X

     of m .

    Of these resu lts, Theorem D6* is contained as a special case in Corollary

    D5.2.

      Fo r the sake of completeness we include in section 2.5 th e deduction

    of Theo rem

     D7*

     from Sch ur's Theorem D 2. Both D6* and

     D 7*

     are generali-

    zations of Theorem D 3.

    1.6. The theory of systems of permutable Sylow subgroups, developed

    in (10) for the case of soluble groups, extends at once in an appropriate

    form to all groups satisfying  C

    s

    m

    .  This has been rem arked for th e case of

    m-separable groups by Gol'berg (8).

    t This terminology seems preferable to that of Hall and Higman (12). In that

    paper, nr-serial groups as defined above were called -nr-soluble. B ut th e m -subgroups

    of m -serial groups need no t always be soluble, whereas all m-su bgroups of a itr-separ-

    able group are soluble, and in particular, all m-subgroups of a m-soluble group are

    soluble. Owing to Burnside's theorem, no distinction arises between -nr-serial and

    -nr-soluble groups as denned here unless -nr conta ins at least three prim es. I hope

    that the present choice of terms is reasonably consistent with Cunihin's.

    5388.3.6 XJ

  • 8/18/2019 Theorems Like Sylow's (Hall)

    5/19

    290

      P . HA LL

    A set of Sylow subgroups

      P

    i

     of the finite group

      G,

     one for each p rime

      p

    i

    in

     w,

     which are permutable in pairs (i.e. such that

      P

    t

    Pj = P

    i

    P

    i

      for all

      i, j),

    will be called a  Sylow m-system  of  G.  Obviously, only a finite number of

    the Pi can differ from 1. Owing to the pe rm utab ility of the P

    i}

     their pro duct

    is an ^-su bg ro up of  G and it is shown in (10), § 2, th at this product m ust

    be soluble. Conversely, a soluble iS^-subgroup H of G has a Sylow m-system

    (whose product is therefore  H  itself) and this will also be necessarily a

    Sylow -nr-system of  G  as defined above.

    In (10), § 4, it was shown that any two Sylow -nr-systems (...,

     P

    i}

    ...)

      and

    (...,

      Qi,...)

     of a soluble -nr-group

     H

      are conjugate in

     H,

      in the obvious sense

    that there is an element

      x

      in

      H,

      independent of

      i,

      such that

      Q

    t

      = P\

      for

    all  i.  Thus we may state at once:

    T H E O R E M A2.  G has at least one Sylow m-system if and on ly if it satisfies

    E^.  G has one and only one class of conjugate Sylow m-systems if and only

    if  it satisfies C^.

    Theorem

      D 3 allows the following conclusions to be draw n.

    (i)

      E% implies E^  for an y subs et  m

    1

      of  to-.

    (ii)

      D^  implies D*

    ni

      for any subset m

    1

      of TO-.

    (i)

      is immediate. As for (ii), let

      G

     satisfy

      D

    s

    m

     and let

      L

      be a maximal

    xo-j-subgroup

      of  G.  Then  L  is contained in. some /S^-subgroup H  of  G.

    B ut

      H  is soluble. Hen ce, by Th eorem D 3, L  is an

     jS^-subgroup

     of H  and

    therefore

      also of

      G.

      S ince t he ^ - s u bg ro u ps of

      H

      are all conjugate in

      H

    and

      t h e ^ - s u b g r o u p s H  of  G are all conjugate in  G, it follows th a t  G has

    only

      a single class of conjugate m axim al

     TOi-subgroups

     and th at these are

    soluble  iS^-subgroups of  G.  Thus  & in fa ct satisfies Z )^ .

    Cunihin

      (2) defines a finite gro up to be

     m-Sylow-regular

     if it satisfies

     C^

    JTl

    for

      every subset

     m

    1

      of

      m.

      Thus we conclude th at

     all groups satisfying D^

    are m -Sylow-regular; an d further, from Th eorem

     A2,

     all such groups possess

    a single class of  conjugate  Sylow m

    x

    -systems for any given subset w

    1

      of

     xtr.

    In particular, this is the case for all groups  G satisfying the hy pothesis of

    Theorem D5 or Corollary D5.1. This statement is a generalization of

    Gol'berg's result about -nr-separable groups. It is obvious that in a

    itr-Sylow-regular grou p, the Sylow T&i-systems are all to be ob tained from

    the Sylow xo--systems by deleting the appropriate terms.

    It must be noted, however, that a group may very well satisfy C^ for

    some suitable

      -m

      w itho ut being -nr-Sylow-regular in C unih in's sense. Fo r

    example, the simple group

      L

    83

      has a single class of conjugate subgroups

    of order 84. These subgroups are dihedral, so that  L

    83

      satisfies Ci^j. But

    L

    83

      has two distinct classes of conjugate subgroups of order 12 and there-

    fore does not satisfy C

    23

    . Thus

      L

    83

      is not 2,3,7-Sylow-regular.

  • 8/18/2019 Theorems Like Sylow's (Hall)

    6/19

    THEO REM S LIK E SYLOW 'S 291

    1.7.  E-Theorems.  Ev ery Z)-theorem includes a C-theorem and a n

    jEJ-theorem by im plication. But , of course, an ^/-theorem can som etimes be

    proved from hypotheses too weak to yield the corresponding D-result.

    The basic ^-theorem is Schur's,

    THEOREM   E l .  If G has a normal S

    m

    ,-subgroup, then G satisfies E

    m

    .

    This may be expressed in a more general form as follows:

    THEOREM E l . *  If K is a normal subgroup o f G such that G/K is a m-group,

    then G has a m-subgroup L such that KL   =  G and K  n  L is nilpotent.

    From this we may deduce the following result, suggested by Cunihin's

    theorems on the factorization of -nr-separable groups (4).

    THEOREM   A3.  Let  m

    0

    ,

      in

    lt

      and vr

    2

     be mutually exclusive sets of primes.

    If  every c.f. of the inite group G is either a

     tcr

    0

     m^group  or else a

     vr

    0

     m

    2

    -group,

    then G

     =

      HK, where H is a  -m

    0

     ^-subgroup, K is a w

    0

     m

    2

    -subgroup,

     and

    H  n  K is a soluble  m

    0

    -subgroup.

    An immediate consequence of Theorem El is the well-known

    THEOREM  E 2 .

      If K is a normal subgroup of G such that K satisfies G

    m

    and G/K satisfies E

    m

    , then G satisfies E

    m

    .

    I t is no t sufficient in this theorem to assume th a t both

     K

      and

      G/K

      satisfy

    E

    m

    .  For example, the group of automorph isms L, of orde r 336, of the simple

    group L = L

    7

    ,  of orde r 168, does not satisfy  E

    23

    ,  although both  L  and  L/L

    do so; because the outer automorphisms of  L  interchange the two classes

    of conjugate octahedral subgroups in  L.

    It will be convenient to mention here the following corollaries of

    Theorem E2.

    COROLLARY  E2 .1 . / / all the c.f. of G satisfy C

    m

    , then G satisfies E

    m

    .

    COROLLARY

      E2.2. ' / /

      the c.f. of G are either m-groups or else m'p-groups

    for various primes p in w, then G satisfies E

    m

    .

    COROLLARY  E2 .3 . / /  G is in-serial and m

    x

      = m, m', mq,  TJJ'P,  or pq,

    where  p is in w and q in  -&•',  then G satisfies E

    mi

    .

    I t w as shown in (11) th at , if a finite grou p G satisfies E

    p

    . for all primes  p,

    then

      G

     is soluble. More generally, by a ve ry similar argum ent, it is easy

    to see that,  if G satisfies E

    m

      and also E

    p

    . for all primes p in  -nr,  then G is

    m-separable.

      It may well be conjectured that

      G

      is soluble whenever it

    satisfies

      E

    pq

      for all pairs of prim es

     p

      and

      q

     which divide its order. Some

    light may perhaps be thrown on this question by studying the behaviour

    of the better known insoluble groups in relation to the propositions  E

    p q

    .

    Here we only consider the symmetric groups and prove:

    THEOREM A4.  Let S

    n

     be the symmetric group of order

     n\

     and letp 

  • 8/18/2019 Theorems Like Sylow's (Hall)

    7/19

    292 P. H A L L

    where p and q are primes. Then  S

    n

      satisfies E

    lhq

      only ivhen p =  2,  q =  3 ,

    and n

      = 3, 4, 5, 7,

     and

      8.

      For n = 5,1, and

      8, E

    u

      satisfies

      C|,3

     but not

      D

    2 3

    .

    Taken in conjunction with Theorem D3, this result shows that, if

      m

    involves more than two primes not exceeding

      n,

      an ^-subgroup of S

    ?l

    ,

    supposing one to exist, can only be insoluble. An exam ple would be the

    £y-subgroups of S

    p

    , where  p  is a prime greater than 5. These are iso-

    morphic with  ^

    p

    -

    x

    .

    1.8.  C-Theorems.  In a long series of papers (see particularly 2 , 3 ,5 , 6 , 7 ) ,

    Cunihin proves a num ber of interesting (7-theorems, of which we m ention

    two,

      deducible from Zassenhaus's Theorems D6 and D7, respectively.

    Let  C%,  be the following proposition about a finite group  G.

    C^: G satisfies  C

    m

     and  NJT  is soluble, where  T  i s any ^-subgroup of

    G   and  N  is its normalizer in  G.

    Then the theorems in question are as follows.

    T H E O R E M

      C l .

      If K is a normal subgroup of G such that K satisfies C

    m

    and G/K satisfies C

    s

    m

    , then G satisfies C

    m

    .

    T H E O R E M  C 2.  If K is a normal subgroup of G such that K satisfies C^

    and G/K satisfies C

    m

    , then G satisfies C

    m

    .

    I t is a corollary of Theorem Cl t hat , if G has a series whose factor groups

    all satisfy  C

    s

    m

    ,  then  G  itself satisfies  C

    s

    m

    .  It is a corollary of Theorem C2

    that, if  G  has a series whose factor groups all satisfy  C

    s

    m

    ,  then  G  itself

    satisfies  C^.  In view of Corollary E2 .1, we can also say th at , if  G has a

    series whose factor groups all satisfy  G

    m

    ,  and if all m-subgroups of  G a re

    soluble, then  G will satisfy

      C%\

      in view of Theorem El, we can add that,

    if

      G

     has a series whose factor groups all satisfy

      C

    m

     and if all xu'-subgroups

    of

      G

     are soluble, then

      G

     will satisfy C^..

    It is natural to make the

    C O N J E C T U R E

      C 3 .

      If K is a normal subgroup of G such that both K and

    G/K satisfy C

    m

    , then G satisfies C^.

    The argument which yields Theorems Cl and C2 from Theorems D6

    and D7 shows that this conjecture is equivalent to the special case already

    mentioned by Zassenhaus, viz.

    C O N J E C T U R E  C 3 . * / /  G  has a normal S

    m

    .-subgroup K, then  G satisfies C

    m

    .

    Zassenhaus states, in (15), p. 126, th at E . W itt has shown how to deduce

    C3*

     from the still more special case where K is a simple group of composite

    order and the centralizer of

     K

      in G is 1. We give here a proof of this red uc-

    tion in a somewhat more precise form.

    Let us say that a group  G involves  an abstract group  V  if there is a

  • 8/18/2019 Theorems Like Sylow's (Hall)

    8/19

    THEOREMS LIKE SYLOW'S 293

    subgroup

      H

      of

      G

      and a normal subgroup

      K

      of

      H

      such that

      H/K

      ~ F.

    If in addition

      (T) < (G), we

      shall say that

      G involves V properly.

    We define a finite group  G to be

     m-exceptional

     if it satisfies the following

    three conditions:

    (i)

      G

     has a normal ^/- subgroup ,

    (ii)

      G

     does not satisfy

      C

    m

    .

    (iii) Every group F properly involved in  G satisfies  C

    m

    .

    It is clear that, if

      G

      satisfies (i) and involves F, then F will also satisfy

    (i).

      Hence every group which satisfies both (i) and (ii) involves a zcr-excep-

    tional group. Thus Conjectures C3 and C3* are equivalent to the state-

    ment that xtr-exceptional groups do not exist. We prove

    T H E O R E M  A 5 .

      Let

      G be m-exceptional. Then both the normal S

    m

    .-subgroup

    K  of G and  also  G/K are simple groups  of  composite order and the  centralizer

    of

      K in G is I, so

     that

      G may

     be regarded

     as a

     subgroup

      of

      the group

      of

      auto-

    morphisms  of K. For any  given prime divisor  q of (K),  there exist S

    m

    -sub-

    groups

      of G

     which

     do not

     leave invariant

      any

      Sylow q-subgroup

      of K. G

      satis-

    fies  C

    mg

     but not D

    mq

    .

    A closer analysis of the properties which a hypothetical -nr-exceptional

    group

      G

     must have, particularly those resulting from (iii) above, would no

    doubt be of some interest. Here we merely stat e:

    THEOREM  C4.

      Let K be normal in G and suppose that both K and G/K

    satisfy C

    m

    . Let T be an S

    m

    -subgroup of K, M its normalizer in G and

    N = KnM. Then M/N  ~  G/K. If S/N is an S

    m

    -subgroup of M/N, then

    either G satisfies C

    m

     or else S/T involves a m-exceptional group.

    In view of Theorem A5, this result contains both Theorems Cl and C2

    as special cases. I t is a corollary of Theorem C4 th at, if

      G

      has a series

    whose factor groups all satisfy

      C

    m

    ,

      then either

      G

     satisfies

      C

    m

      or else

      G

    involves a -or-exceptional group.

    As Zassenhaus remarks (loc. cit.) the Conjectures C3 and C3* would

    follow from either of the long-standing conjectures:

    (a)

      All groups of odd order are soluble.

    (6) The group of outer automorphisms of a finite simple group is soluble.

    They would also follow, as is shown by Theorem A5, from the conjecturef

    (c) If A is a subgroup of the group of automorphisms of the finite group G

    and if

     (A)

     and

      (G )

     are coprime, then for every prime

     q

     dividing

      (G),

     A

      leaves

    invariant some Sylow ^-subgroup of

     G.

    That (c) is true if either

     A

      or  G is soluble follows easily from Zassenhaus's

    f That the conjecture  C3* is equivalent to (c) is proved by D. G. Higman, Pacific

    J.

      Math. 4 (1954), 545-55.  I am indebted to the referee  for  this reference.

  • 8/18/2019 Theorems Like Sylow's (Hall)

    9/19

    294  P .  HALL

    theorems

     D6 and

     D 7.

      I n

     this connexion

     it may be

     of some intere st

     to

     point

    ou t

      th e

      following consequence

     of

     Theorem

     D 5.

    T H E O R E M

      A6.

      If 0 is a  finite group satisfying  E^ and A is a soluble

    m-subgroup of the group of automorphisms  of G,  then A  leaves invariant some

    S

    m

    -subgroup  of G.

    For we can take  G  in its regular representation and regard  AG  as a

    subgroup of the holomorph of  G.  By Theorem D5, A G  satisfies  D

    m>

     since

    AG/G  r^j A  and so satisfies D^,  while G satisfies

      E%.

      Hence A  is contained

    in some #

    OT

    -subgroup HoiAG.  Therefore A leaves invariant the ^- su bg ro up

    H  n  G of G.

    We remark tha t, if A  is a soluble subg roup of the group of autom orphism s

    of the arbitrary finite group  G  and if there is at most one prime  q which

    divides both  (A )  and  (G), then we can take  vr in Theorem A6 to consist of

    the prime divisors of (^4) together with

      q

     if necessary. In this case then ,

    A

      leaves invariant some Sylow ^-subgroup of

      G,

      no special assumption

    about

      G

     being required.

    W e hav e already n oted (Corollary E2.3) th at m-serial groups satisfy

      E

    m

    .

    I t w ould follow from Conjecture C3* th a t they even satisfy D

    w

    . In fact

    we have:

    T H E O R E M  D 8.  Let G be a nr-serial

     group.

      If L is a  soluble  m-subgroup

    of  G and if H is an  S

    m

    -subgroup  of G, then L 

  • 8/18/2019 Theorems Like Sylow's (Hall)

    10/19

    THEO REM S LIK E SYLOW 'S 295

    We may therefore suppose that  K  is nilpotent and have only to prove

    the existence of a -nr-subgroup   S  such that  KS  =  G.  If there is a normal

    subgroup

     K

    x

     of G such tha t 1 <

      K

    x

      <

      K,

      we apply the induction h ypothesis

    to  G/K

    x

      and obtain a subgroup  S

    x

      ^  K

    x

      such that  S

    x

    /K

    x

      is a tr-group and

    KS

    X

      = G.  Then we apply the induction hypothesis to  S

    x

      and obtain a

    -nr-subgroup  S  such that  K

    x

     S  =  S

    x

    .  This gives KS = KK

    X

     S  =  KS

    X

      =  G

    as required.

    Finally, let  K he a,  minimal normal subgroup of  G.  Being nilpotent, it

    is an Abelian #>-group. If

      p

      is in -nr, we can take

      S

      =

      G.

      If

      p

      is in -a/,

    then K  is a norm al Abelian /S^-subgroup of G. B y Theorem D2, G satisfies

    E

    m

     and we can take for

      S

      any /S^-subgroup of

     G.

    Proof of

     A3.  We begin with an obvious rem ark. Let

    G=G

    0

    >G

    x

    >...>G

    r

    =l

    be any series of subgroups of  G.  Let  H  and  K  be subgroups of  G  such

    Then

    hat , for each

      i

      = 1,2

    G = HK.

    For, by hypothesis,

    ,...,

      r,

      e i ther G

    e i the r  G

    r

    _

    x

      ^

    r

    t

    _

    x

      0, that

      G

    t

     ^ J?iT. If G ^ ^ ^JEf,

    then also G^-  ̂ < ^ ^ , since ^ is normal in

      G^^

      Hence

    O

    t

    _

    x

     

      G

    x

     > ... >

      G

    r

     =

      1 be a chief series of

      G,

      so that each

     G

    is norm al in

     G.

     Ev ery factor group G^JGi  is a direct produ ct of isomorphic

    c.f. of

      G.

      Hence we may divide these

      r

      factor groups into two mutually

    exclusive classes, putting in the first class those which are   TO-

    0

     xo-j-groups

    (including those, if any, which are xn-

    0

    -groups) and in the second class all

    the rest. We now show th at th ere is a  m

    0

     to^-subgroup  H  and a  TO-

    0

    'OT

    2

    -

    subgroup

     K

      in G such tha t (i) if

     G

    i

    _

    x

    lG

    i

     is in the first class, the n G

    {

    _

    x

      ^  G

    i

     H,

    while  K^JKi  is nilp ote nt, and (ii) if  G

    i

    _

    x

    IG

    i

     is in th e second class, th en

    G^

    x

      ^

      G

    t

    K

      and

      H^JHt

      is nilpotent. Here we have written

      H

    i

    =  G

    i

    C\ H

    and

     K

    i

    =

      6^ n

     K.

      Arguing by induction on r, we m ay assume the ex istence

    of subgroups

      H

      and

      K

      containing

      G

    r

    _

    x

      and such that

      H/G

    r

    _

    x

      is a  vr

    0

     m

    x

    -

    group,  K/G

    r

    _

    x

      is a

      vr

    0

     nr

    2

    -group and (i)* if  G^JGj  is in the first class and

    i

      <

      r,

     then

      G

    t

    _

    x

     <

      G^

      while

     K

    i

    _

    x

    IK

    i

      is nilp ote nt, (ii)* if G^JGi  is in the

    second class and  i 

  • 8/18/2019 Theorems Like Sylow's (Hall)

    11/19

    296 P. HALL

    now

      G

    r

    _

    1

    /G

    r

     ~ G

    r

    _

    x

     is in the first class, we take H  =  H,  while for  K  we

    take a  vr

    Q

    -nrg-subgroup of  K  such that

      G

    r

    _

    x

    K

     = K  and  G

    r

    _

    x

     n K  is nil-

    potent. Such a subgroup K  exists by Theorem El*. Then it is clear from

    (i)*

     and (ii)* that

     H

      and

     K

      satisfy (i) and (ii) for each

     i =

      1,

     2,..., r.

      Also

    H

      =

      H

      is a  xtr

    0

     TDygroup.

      Similarly if

      G

    r

    _

    x

    /G

    r

     is in the second class.

    The remark made above shows that

      G = HK.

      The intersection

    is a TD-

    0

    -group since  vr

    Q

    , m

    v

      and  vr

    2

     are mutual ly exclusive. If  D

    i

      = G

    t

    C\ D,

    then

     D

    i

    _

    1

    ID

    i

     ~  G

    i

     Di-JG^ which is a subgroup of both G

    i

    H

    i

    _

    1

    IG

    i

     ~ H^JH^

    and of G

    i

     K^JGi  ~ K^JK^  But for each i, one of the two groups H^JHi  and

    Ki.JKj  is nilpotent. Hence all the factor groups  D

    i

    _

    1

    /D

    i

      are nilpotent

    and D  is soluble. Thus A3 is completely proved.

    2 . 3 .  LEMMA  1.

      Let K be a normal subgroup of G and let H be an S

    m

    -

    subgroup of G. Then K

      n

      H is an S

    m

    -subgroup of K and KH/K is an

    S

    m

    -subgroup of G/K.

    For let  (G:K) = h and  (K) = k.  Then

    {G)

    m

     =  h

    m

    k

    m

      = (H) = (H:Kn H)(K

      n

      H).

    But

     H/K

      n

     H ~ KH/K,

      which is a w-subgroup of

     G/K,

     so that

      (H:Kn H)

    divides h

    m

    \  and K  n H  is a To--subgroup of K  so that  (K n H)  divides  k

    m

    .

    Hence  (KH:K) = h

    m

      and  (K n H) — k

    m

    , as required.

    Proof of E2. Let  K  satisfy  C

    m

     and let 6?/Z satisfy  E

    m

    .  Let T be an

    /S^-subgroup of

     K

      and let

     N

     be its normalizer in

      G.

      For any

     x e G, T

    x

      is

    an /S^-subgroup of iT and hence is conjugate to  T  in K  so that  T

    x

      = T

    v

    with y e . Thus

     xy-

    1

      e N

      and

     NK = G.

      Hence JV/Z n

     N

      ~

      G/JK"  SO

     that

    JV/X n JV has an #

    OT

    -subgroup H/K  n JV. Apply Theorem El to the group

    H/T  with the normal ̂ --subgroup K  n

     iV/2

    7

     and we obtain an ^-subgroup

    U/T  of H/T.  Then  (U:T) = {H.K D N) = h

    m

      and (T) =  /c

    OT5

     so that  U

    is an /S^-subgroup of G.

    L E M M A

      2. .Le£

     H = H

    x

    xH

    2

    x

      ...xH

    r

    .

      If each H

    t

      has a given one of the

    properties E

    w

    , C

    w

    , D

    m

    , E*

    m

    , C

    s

    m

    , D

    s

    m

    , E , or C^, then H has the same property.

    The proof is straightforward and may be omit ted .

    Suppose now that all the c.f. of  G  satisfy  C

    m

    ,  and let  K  be a minimal

    normal subgroup of  G.  Then  K  is a direct product of c.f. of  G.  Hence  K

    satisfies  C

    m

      by Lemma 2. Corollary E2.1 now follows from Theorem E2

    by an induction argument on (

    0),

     since we may assume tha t

      G/K

      satisfies

      E

    m

    .

    Since trr-groups satisfy  C

    m

      trivially and

      TO-'^-groups

     satisfy  C

    m

      by Sylow's

    theorem, we obtain Corollary E2.2 as a particular case of E2.1.

    As for Corollary E2.3, this follows from E2.2 for

      m

    1

      = w, m', vrq,

      and

  • 8/18/2019 Theorems Like Sylow's (Hall)

    12/19

  • 8/18/2019 Theorems Like Sylow's (Hall)

    13/19

    298  P .  HALL

    We make  a few pre lim inar y simplifications. Since  G/K  satisfies  D

    m

    , we

    have  L 

  • 8/18/2019 Theorems Like Sylow's (Hall)

    14/19

    THEOREMS LIKE SYLOW'S 299

    Now let

      Q

      be the normalizer of

      P

      in

      G

      and let

      Q

     =

      K

      n

      Q.

      We first

    prove (5) that  Q satisfies El^.  Since  P  and  Q  are both normal in  Q and

    Pn Q ^. Lr\ K = I,

      we have

      PQ = PxQ.

      Let the distinct primes

    dividing  (G)^  be p = p

    x

    ,  p

    2

    ,...,  p

    r

      and let  Q

    i

     be a Sylow

    r

    subgroup of Q

    for

      i —

     1, 2,..., r. Since T satisfies Z^ we may choose an iS^-subgroup T$

    of  T  such tha t P x (^ < T,. We note that ^ = Z n 2J is an ^ -s ub gro up

    of̂ fiT by Lemma land

      is

     therefore nilpotent. Also

     T

      = i£ P and

      so 7^

      =

      S

    i

    P.

    Hence  T

    t

      contains a unique Sylow ^-subgroup P^ containing  P  and the

    centralizer

      Z

    t

      of

     P

    t

     in

     T

    t

     is nilpotent and contains

      Q^

      Since

     T

      satisfies

      D

    mi

    the subgroups  T̂  are all isomorphic. More precisely, there is an isomorphism

    mapping

     T

    t

     on to

     T

    x

     and at the same time mapping

     P

    t

     on to

     P

    x

     and therefore

    Zi onto Z

    x

    .  If this isomorphism maps Qi onto ii^for i =  2,

     3,...,

     r, itfollows

    that iẐ <

      Z

    x

    .

      Since

      Z

    x

     is nilpotent, the product

      R

      of

      R

    x

      = Q

    x

    ,

      R

    2

    ,...,

      R

    r

    is direct; and since  P  -group,

     P

      < P

    x

     would imply that

    P < $ n

     P

    lt

      so that

      P = LOT

      would not be an ^- su bg ro up of

      Q

     n T

    7

    and therefore, by Lemma 1, L would not be an iS^-subgroup of  Q.  Hence

    P =

      P

    x

     is a Sylow ^-subgroup of

     H

      n T. Also

      H 0 T is

     normal in J?. By

    Sylow's theorem, the normalizer  Q C\ H = U of P in H  satisfies  H ^. TU.

    Since T =

      KP

      and P <

      U,

     we even have

      H ^KU.

      Since

     KH = G by

    (2),  we have i£?7 = £. But  LnK =  1. Hence (17) > (X). But £ is an

    /S^-subgroup of

      Q

     and

      ?7

     is a -or-subgroup of

      Q.

      Hence

      U

     is an /S^-subgroup

    of  Q.  But  Q  satisfies  D

    m

      by induction. Hence  L = U

    x

      ^ ^

    x

      for some

    x e G

     and Theorem D5 is completely proved.

    Corollary  D5.1  follows immediately. If  G has a series all of whose factor

    groups satisfy

      E^,

      then all the c.f. of

      G

     satisfy  E

    1

     ^ by Lemma 1. If

      K

      is

    a minimal normal subgroup of  G, then  K  satisfies  E

    1

    ^.  by Lemma 2. The

    c.f. of

      G/K

      all satisfy

      E%

     and we may suppose inductively that

      GjK

    satisfies

      D

    s

    m

    .

      Then

      G

     satisfies

      D%.

     by Theorem D5.

    Corollary D5.2 is a particular case of  D5.1.  For if

      G

     is -nr-separable and

    vr

    1

     is any non-empty subset of

      w,

      then every c.f. of

      G

     is a

      w'

    x

    p

    -group for

    some prime

      p

      in

      -m

    x

     and therefore satisfies

      E^

      by Sylow's theorem.

    Corollary D5.3 is a special case of D5.2.

    Proof of A1. Suppose H  and K  are S

    m

     -subgroups of  G and that both  H

    and

     K

     have Sylow series of complexion

      (p

    1

    ,P2>-'->P

    r

    )>

     where

     p

    x

    ,...,  p

    r

     are the

    distinct primes dividing  (G)

    m

    .  If  H

    x

      and  K

    x

      are the normal

      S

    VlPi

      Prl

    -

    subgroups of

     H

      and

      K,

      respectively, then

      H

    x

     and

      K

    x

      have Sylow series of

  • 8/18/2019 Theorems Like Sylow's (Hall)

    15/19

    300 P. HA LL

    complexion (Pi,.--,Pr-i)

      a n

    d

      a r e

      $p

    lf

    2>,

      3

    »

    r

    _

    1

    -subgroups of

      G.

      Arguing by

    induction on r, we may suppose

      H

    x

      and

      K

    x

      are conjugate in

      G.

      We may

    then take

     H

    x

      = K

    x

    ,

     replacing

     K

      by a conjugate if necessary.

      H/H

    x

    smd.K/H

    x

    are then two Sylow ̂ -su bg rou ps of

     N /H

    x

    ,

      where

     N

      is the normalizer of

     H

    x

    in

      G.

      Hence

      H

      is conjugate to

      K

      in

      N.

    2.5.  Proof of

     D7*.  I t will be sufficient t o show th a t a ttr'-soluble grou p

     G

    satisfies

      D

    m

    .

    Let

      K

      be a minimal normal subgroup of G.  Then

     K

      is either a -nr-group

    or else an Abelian g-group for some prime

     q

     in

      m'.

      Also

      G/K

      is icr'-soluble.

    Arguing by induction on

      (G),

     we may assume th at

      G/K

      satisfies

      D

    m

    .

      If

      K

    is a -nx-group, then

      G

     satisfies

      D

    m

      by Lemm a 4. Thus we may assume t ha t

    K

      is an Abelian -nr'-group.

      G

     satisfies

      E

    w

      by Theorem E2. Let

      H

      be an

    ^ - s u b g r o u p o f

      G

     and

      L

      any xtr-subgroup of

     G.

      Since

      G/K

      satisfies

      D

    m

      by

    the induction hypothesis, we have  L 

  • 8/18/2019 Theorems Like Sylow's (Hall)

    16/19

    THEOREMS LIKE SYLOW'S 301

    K  cannot be Abelian or  G  would satisfy  C

    m

      by Theorem D2. Hence

    K

      =

      K

    1

    xK

    2

    X...xK

    r

    ,

      where the

     K

     are isomorphic simple groups of com-

    posite order. The

      r

      subgroups iQ are the only normal subgroups of

      K

    which are simple. On transforming  K  by the elements of H,  the  K

    t

      must

    therefore be permuted transitively among themselves. Hence if  N  is the

    normalizer of

     K

    x

      in

      G,

     we have

     K

      <

      N

      and

      (G:N) = r.

      Let

      q

     be a prime

    divisor of  (Kj)  and let  Q

    x

      be any Sylow (/-subgroup of K

    x

    .  Let  N  be the

    normalizer of  Q

    x

      in  N.  By Sylow's theorem,  K

    X

    N  =  N  and therefore

    KN = N.  Hence  N/K  n N  ~ F / Z and, by Theorem El , 2V contains an

    £

    OT

    -subgroup

      U,

     so that

      {K n N)U = N

      and therefore

      KU = N.

      Thus £7

    is an iS

    m

    -subgroup of  N.  Since 6r = iT£T =  KH

    X

      and  K ^. N,  the inter-

    sections  V = H  r\ N  and 1̂ = T̂

    x

     n iV are also ^. -subgroups of  N.

    Suppose if possible that

      r

      > 1. Then

      N < G

     and hence

     N

      satisfies

      C

    m

    .

    Consequently

      U, V, V

    x

     are conjugate in

      N.

      Replacing / / and

      H

    x

      by con-

    jugates, we may assume tha t  U =  V = V

    x

    . We then have  (H:U)  =  r and

    may choose the elements  h

    x

      =  1, h

    2

    ,..., h

    r

     in H  so that

    . £*

     = jfî (i = i,  2,..., r).

    Then Qi* =  Qi is a Sylow ^-subgroup of

     iQ

     and hence

      Q = Q

    1

    xQ

    2

    X...xQ

    r

    is a Sylow ^-subgroup of  K.  For any  xe H,  we have  ^x  = u^-, where

    w

     e

      C7

     and j = j (i ,3 ). Hence Qf =

      Q^

    x

      = Q * = Q$ = Q

    p

      since

      U

     <

      N.

    For a fixed  xe H, i -+j(i,x)  is a permutation of 1, 2,..., r. Hence  Q

    x

      = Q

    and so i/ is contained in the normalizer  R  of  Q in 6r. Similarly  H

    x

     is con-

    tained in R.  Since $ is not normal in K, we have R < G and so 72 satisfies

    C

    m

    .  Hence the ^-s ubgro ups  H and H

    x

     of i? are conjugate, a contradiction.

    Thus  r  = 1 and i£ is a simple -nr'-group of composite order.

    By Sylow's theorem

     KR = G,

     so tha t the ^-subgroups of

     R,

      which are

    all conjugate in  R  as we have already remarked, are also ^- su bgr ou ps

    of

      G.

      There is therefore a uniquely determined class of conjugate

    ^-sub group s of G associated with any given prime divisor q of (K), viz. those

    which belong to the normalizer in

     G

     of some Sylow (/-subgroup of

     K

    . Since

    G  does not satisfy  C

    m

    ,  there must be some  S

    m

     -subgroups of  G  which do

    not leave invariant any Sylow (/-subgroup of

     K.

      If

     H

     is one of these, then

    H  cannot be contained in any

    m(2

    -subgroup  L  of  G, for  L  n K  is normal

    in

      L

      and is a Sylow (/-subgroup of

      K.

      Thus

      G

      cannot satisfy

      D

    mq

    .

      But

    G satisfies  E

    mq

    bj  Corollary D5.3, or more simply by Sylow's theorem and

    Theorem El. In fact,  G  satisfies C

    OT(r

      For let  L  and  L

    x

      be any two

    /S^-subgroups of G. Then LP\K  and L

    X

    D K are Sylow (/-subgroups of K  and

    therefore conjugate. Replacing  L

    x

      by a conjugate if necessary, we may

    suppose that  Lc\ K = L

    x

    n K  =  Q,  so that both  L  and  L

    x

      belong to the

    normalizer  R  of  Q. L/Q  and  LJQ  are then ^-s ubgrou ps of  R/Q  and

  • 8/18/2019 Theorems Like Sylow's (Hall)

    17/19

    302 P. H A L L

    therefore conjugate because  R/Q  is properly involved in  G and so satisfies

    C

    m

    .

      Hence

      L

      and

      L

    x

      are conjugate.

    Let  Z be the centralizer of K  in

     G.

     Then Z is norm al in 0  and K  n  Z  = 1

    because if is simple bu t not cyclic. Since  G/K  is also simple,  Z  ^

      1

     would

    imply that  G =  KxZ.  This would make Z a normal and therefore the

    unique ^-subgroup of  G.  This is impossible since  G does not satisfy  C

    w

    .

    Hence

      Z

      = 1 and Theorem A5 is completely proved .

    Proof  of C4. Let  K  be a normal subgroup of  G  such that both  K  and

    #/ £ satisfy C^. By Theorem E2,

      G

     satisfies

      E

    m

    .

      Suppose th at  G does not

    satisfy

      C

    m

      and let

      H

      and

      H

    x

      be

      /S .̂

     -subgroups of

      G

      which are not con-

    jugate in  G.  The ^ - subgroups  H  n K  and H

    X

    C\  K of K  are conjugate and

    we may assume  H f) K =  H

    X

    C\ K = T,  replacing  H

    x

      by a conjugate if

    necessary. Then both

      H

      and

      H

    x

      are contained in the normalizer

      M

      of

      T

    in   G.  Let  N = K n M.  Since Z satisfies C

    OT

    , we have  KM = G as in the

    proof of Theorem E2 . Hence   M/N  ~ £ / Z and satisfies  C

    w

    :  The

    ^ - subg roups  NH/N  and  NHJN  of Jf/iV are therefore conjuga te and we

    may assume that

      NH = NH

    X

      =

      S,

      replacing

      H

    x

      by a conjugate if neces-

    sary.  S/T  has the normal ̂ - su b gr ou p N/T  and has the two non -conjugate

    S^-subgroups

      H/T

      and

      HJT.

      Hence

      S/T

      does no t satisfy

      C

    m

    .

      Therefore

    S/T

      involves a w-exceptional group and Theorem C4 is prove d.

    2.7.  Proof of  D8. We argue by induction on  (G).  Let  K  be a minimal

    normal subgroup of

      G, H

      any /8^-subgroup of

      G,

     and

      L

      any xo--subgroup

    of  G.  We have to prove that, if either  L  is soluble or if  G  involves no

    •nr-exceptional group, then

      L

      ^

      H

    x

      for some

      x e G.

      Since

      KL/K

      is a

    -nr-subgroup and  KH/K  is an /S^-subgroup of  G/K,  we may suppose by

    induction that

      L

      <

      KH

    X

      for some

      x

     e

      G;

     for if

      L

      is soluble, so is

      KL/K,

    while if  (3 involves no xir-exceptional group, neither does  G/K.  Also,  KH

    X

    is a -nr-serial group and involves no -nr-exceptional group unless G involves

    one.

      The result now follows by induction if  KH

    X

     

  • 8/18/2019 Theorems Like Sylow's (Hall)

    18/19

    T H E O R E M S L I K E S Y L O W ' S

      303

    2.8.  Proof

     of A4. Let p < q

     <

      w, where p

      and

     q

     are

     primes,

     and let

     H

    be an $

    pg

    -subgroup of the sym metric group S

    n

    . By a theorem of Burn side,

    H is

     soluble.

    Suppose first that  H is a  primitive permutation group. Then n = r

    m

    ,

    where r is

     a

     prime, an d r is either p

     or q.

      Also, H is contained

     in

     the holo-

    morph  H of the  elementary Abelian group  of order  r

    m

    .  Cf.  Speiser (13,

    Satz 98). Since

     (H)

    r

     = r^-

    m(

    -

    m+x

    >

     and

      (H)

    r

     = ( SJ

    r

     =

      r

    i+r+r*+...+r»-»

    } w e  m u s t

    h a v e

      £

    This equation implies that either m =

      1

     or else

     m =

      2,

     r =

      2. In the case

    m

     =

      1,

     if is the

     metacyclic group

     of

     order r(r—1),

     so

     tha t

      we

     must have

    r =

     q = n =

      l-\-kp.  But then  (?>

    n

    )

    p

      ^

      p

    k

    ,

     so  t ha t

      &

     must be divisible by

    p*-

    1

    .

      Thus either k =

      2,

     p =

      2, or

     k.=

      1. But

     (S

    5

    )

    2

     = 8 so

      tha t  &

     =

      2

     is

    impossible. Hence jfc=l,^>

      = 2, g = 3, and

     H =

     S

    3

    .

    If, on  the other hand, m = 2, so th at r =  2 =  p, we have w = 4 and so

    q = 3, H =  S

    4

    . We conclude tha t the only cases where H can be primitive

    are when p =  2, q = 3, and w = 3 or 4, and in these cases H is  the whole

    of S,

    /r

      In all other cases

     H

     is  either imprimitive or  intransitive.

    Ne xt suppose th at H is intran sitive , with com ponents of degrees

     I

     and m

    where £+ra

     =

      n.

      Then

     i/ is

     contained

     in S

    z

    x

     S.

    m

     and, since

     it

     must be

     an

    $p

    S

    -subgroup

      of

      this group, both

      £j and S

    m

      must satisfy

      JE^.

      Further,

    # cannot exceed both I and m, since the n  (H ) would be prime to q, contrary

    I

     of

      S; X S

    wl

     in S

    M

     must

      be

     prime

      to

     pq,

    mj

    since otherwise H  could

     not be an

     $p

    g

    -subgroup

     of

     S

    M

    .

    A similar argument applies

      if

      JÊ

     is

      imprimitive, with

      e

      systems

      of

    imprimitivity  of d  symbols each. Then  n = de and H  is  contained  in

    S^ISg,

      a

      maximal imprimitive subgroup

      of S

    M

      corresponding

      to

      this

    factorization  of n.  The order  of  2^1; £

    e

     is  (d\)

    e

    e\ and therefore  q  cannot

    exceed both d

     and e.

      Since  G

     =

      S

    d

    1 S

    e

     has a

      normal subgroup K  which

    is the direct prod uct  of e  factors isomorphic with  2

    d

     and G/K ~ S

    e>

     it is

    clear from Lem ma 1 th at bo th

      2

    d

     and 2

    e

      must satisfy  E

    pq

    .  Further

     the

    index n\/(d\)

    e

    e\ of G in S

    n

     must be prime to pq.

    If therefore  JET

     is

      either imprimitive

      or

      intransitive, there

     is

     necessarily

    an integer n

    x

      such that  q 

  • 8/18/2019 Theorems Like Sylow's (Hall)

    19/19

    304 T H E O R E M S L I K E S Y L O W ' S

    A subgroup

      of

      index

      35 in S

    7

      c a n n o t

      be

      t r ans i t ive

      and so

      m u s t h a v e

    the form E

    3

    x S

    4

    . Again

      we

     h a v e

      a

      single class

     of

     #

    2 3

    - subgroups . None

     of

    these con ta ins

      a

      cyc l ic pe rmuta t ion

      of

      order

      6.

      T h u s

      S

    7

     satisfies

      C

    23

     but

    n o t  D

    2

     3

    .

    A subgro up of index 35

     in S

    8

     c a n n o t

     be

     in t ran s i t ive s ince

     all th e

     b inomia l

    coefficients 8, 28, 56 ,70 ar e even .

      I t

     cann ot be pr im it ive s ince th e ho lom orph

    of

      the

      e le m e n ta r y g r o u p

      of

      order

      8 has

     index

     30 in S

    8

    .

      H e n c e

      it

      m u s t

      be

    im p r im i t i v e

      and

     on ly

      the

      subgroups

     of the

      form S

    4

    1 S

    2

      are

     la rge e noug h.

    Aga in

      we

      h a v e

      a

      single class

      of

      conjugate

      S

    2

     3

    -subgi\oups. None

      of

      t h e m

    con ta ins

     a

     s u b g r o u p

     of

     th e form 2

    2

    1 E

    4

    . Thus

      2

    8

      satisfies C

    2 3

     bu t not

     Z )

    23

    .

    I t r ema ins

      to

      show tha t ,

      if n  > 8, E

    n

     does

     not

     satisfy

      E

    2 3

    . Suppose

     the

    c o n t r a r y

      and

      choose

      n  > 8 as

      smal l

      as

      possible

      so

      t h a t

      2

    n

      has an

    S

    2 3

    - subgroup H.  If H  is in t rans i t ive w i th componen ts  of  degrees I and m,

    w h e r e  I 

    d

    %

     S

    e

     where

     de

     =  n.  As we have

    also seen, both  2,

    d

     and 2

    e

     must have $

    23

    -subgroups and  n\/(d\)

    e

    e\  must be

    prime to 6. By our choice of n,  both d and e must be ^ 8, ^ 6, and   ̂ 1.

    But in this case H is transitive so that  n  must be of the form 2

    a

    3^. Thus

    d

     and e can only be 2, 3, 4, or 8. Again it is easy to verify that none of the

    eleven relevant indices

     is

     prime to

     6.

     This concludes the proof of Theorem

     A4.

    REFERENCES

    1.

      W. BURNSIDE, Theory

      of

     groups

     of

     finite order,  2nd edition (Cambridge, 1911).

    2.

      S.

     A.

      CUNIHIN,

     Mat. Sbornik,

     N .S . (25) 67

     (1949), 321-46.

    3.

      Doklady

     Akad.

      Nauk S.S.S.R.

      N.S.

     73

     (1950), 29-32.

    4.  ibid.

     95

     (1954), 725-7.

    5.  ibid.

     66

     (1949), 165-8.

    6. ibid.

     69

     (1949), 735-7.

    7.  Mat. Sbornik, N.S.

     (33) 75

     (1953), 111-32.

    8.

      P.

     A.

     GOL'BERG, Doklady

     Akad.  Nauk S.S.S.R.

      N .S . 64

     (1949), 615-8 .

    9.

      P.

     HAXL,

     J. London Math. Soc. 3

     (1928), 98-105.

    10.  Proc. London Math. Soc. (2) 43

     (1937), 316 -23.

    11.  J . London Math. Soc. 12

     (1937), 198-200.

    12.  and GRAHAM HIGMAN,  Proc. London Math. Soc. (3) 6

     (1956),

     1-42.

    13.

      A.

      SPEISER,

      Theorie der Gruppen

     von

     endlicher  Ordnung,

     3rd

      edition (Berlin,

    1937).

    14.

      H. WIELANDT,

     Math. Zeitschrift,

     60

     (1954), 40 7-8 .

    15.

      H. ZASSENHAUS,

     Lehrbuch der  Gruppentheorie  (Leipzig and Berlin, 1937).

    King's College,

    Cambridge