Supplementary Meimoir on the Theory of Matrices .
By A. CAYLEY , Esq. , F.R. S. Received October 24 , -Read December 7 , 1865 .
M. HERMIT , in a paper " 1 Sir la theory de la transformation des fonctions Abeliennes , " Comptes Rendus , t. xl .
( 1855 ) , pp. 249 , &amp;c. , establishes incidentally the properties of the matrix for the automorphic linear transformation of the bipartite quadric function Xw ' +yz'-zy'-wx ' , or transformation of this function into one of the like form , XW'+YZ'-ZY'-WX ' .
These properties are ( as will .
be .
shown ) deducible from a general formula in my " Memoir on the Autom-orphic Linear Transformation of a Bipartite Quadric Function , " Phil. Trans. vol. exlviii .
( 1858 ) , pp. .
39-46 ; .
but the particular case in question is an extremely interesting one , the theory whereof is worthy of an independent investigation .
For convenience the number of variables is taken to be four ; but it will be at once seen that as well the demonstrations as the results are in fact applicable to any even number whatever of variables .
Article Nos. 1 &amp; 2.-Notation and Remarks .
1 .
I use throughout the notation and formule contained in my " I Memoir on the Theory of Matrices , " Phil. Trans. vol. cxlviii .
( 1858 ) , pp. 17-37 , and in the above-mentioned memoir on the Automorphic Transformation .
With respect to the composition of matrices , the rule of composition is as follows , viz. .
, any line of the compound matrix is obtained by combining the corresponding line of the first or further component matrix with the several columns of the second or nearer component matrix ; it is~very convenient to indicate this by the algorithm , ( a , b , c lg , ,y )-a , c ) 11 a ' , b6 c ' j| 3a,.'I ( a ' , bt c ' ) all , b~l ' " at/ , " it , " y(all " V'l e~u , , , which exhibits very clearly the terms which are to be combined together ; thus in the upper left-hand corner we have ( a , 6 , c(a , a ' , a " ) , and so for the , other places in the compound matrix .
2 .
It is not in the Memoir on Matrices explicitly remarked , but it is easy to see that sums of matrices , all the matrices beingof .
the same order , may be multiplied together by the ordinary rule ; thus ( A+B)(C+D)=AC+AD+BC+BD : this remarlk will be useful in the sequel .
MDcccLxI .
E Article Nos. 3 to 13.-First Investigation .
3 .
We have to consider the formulae for the automorphic linear transformation of the function xw'+yz'-zy'/ -wx , that is , of the function -,.,1 , .
, 1 , .
, .
1 , .
, .
, =(2Qx , y , Z , xwq( , y ' , z ' , w ' ) , viz. , if the variables are transformed by the formula ( X* Z , : , o)-(rIKX , Y.3 ZW )I ( x ' , Z ' &amp; ' ~)=j1~~ , Yt , V , Z[WI ) , then the matrix ( LI ) is such that we have identically PIXx , Y , Z , qu](4 Lie , Z ' , qvf ) =( ( KXI YJ Z~ W3X ' Y1Z ' , W ) the expression for ( LI ) is given in my memoir above referred to ; viz. , observing that the matrix ( 2 ) is skew symmetrical , then ( No. 13 ) we have where T is an arbitrary symmetrical matrix .
4 .
I propose to compare with the matrix II the inverse matrix HI- .
Recollecting that in the theory of matrices ( ABCD)-1=D-'C-1B-1A- ' , we have Il-l Q-21P0+yXQ0 y ) -l0 ; and it is to be shown that II and II-1 are composed of terms which ( except as to their signs ) are the same in each , so that either of these matrices is derivable from the other by a peculiar form of transposition .
It is to be borne in mind throughout that T is symmetrical , Q skew symmetrical .
5 .
I write for greater convenience l ' 21Q(T§ 2)(T2 &gt; '2 , and I compare in the first instance the matrices ( T -9)(T+2 ) ?
' and ( T+2)(XT2)- ' 6 .
Any matrix whatever , and therefore the matrix ( T + Q)-1 , may be exhibited as the sum of a symmetrical matrix and a skew symmetrical matrix ; that is , we may write ( T+2')=T ?
+2 , where T ' is symmetrical , 2 ' is skew symmetrical .
We have then where , here and in what follows , 1 denotes the matrix unity .
Moreover T a2= tr.(T+ ~ and thence ( T2)- ' = ( tr.(T+U2 ) &lt; 1 _tr .
( T+ &amp; )= tr.(T'+Q')=T_ l-i2 that is , ( Y -Q)T/ fQf , and thence also ( T- &amp; ?
)(Tt2 ?
1= ( TQ)(T2 ' ) =1 .
We have therefore ( T Q)(T + )-i = ( T + 12:)(rTf+ t ) = I2 Q(rT + 15+ ( T + 1)(T Q)-i=(T U+ 22)(TT ) 2 1)2u(T'- , Q ' ) .
7 .
Suppose for a moment that T'+2'=( a , 6 , c , d e , f , g , h 5 , I , k , I ml , 0 , _p and therefore rT-'_ a , :ez , tj in b5 f , v , n K , g , k , o d , I , pj 8 .
We have. .
I )( a , b , c , d ). .
1 .
e , f , e g , h 1 ... rn , n , o , p ( a , e , i , in ) , ( b , f j , ) , ( , g k , o ) , ( d , h , p ) ( , -1 .
-e , -.f , -g , h -a , -d6 , - .
-d And similarly , n(v fQ ' )a 1 )( , e , , m ) -1 .
6 , f , , I , d c , g. k , o ( a , b c , d ) , ( e , f , g , h ) , ( i , J k , 1 ) , ( i n o , ) =(-d , -h , -1 , -p ) .
c , -g , -k , -o.,6 , f j , # .
a , e , ; , m 10 .
Hence also T_ )(T + fQy'-l-( l +2n.2n , 20 , 2p ) 2i , 1+2j , 2k , 21.-2e , --2f , 1 -2g -2h -2a 5 -2bl -2c , 12d and ( T+ ?
)(T-2)-'=( 1--2d , -2h , -21 -2p ) -2c , 12g , 2k , -2o 26 , 2f , 1+2j , 2n -2a , 2e , 2i , 1+22m , so that these matrices are composed of terms which , except as to the signs , are the same in each .
11 .
Now in general if c3=( , S3 , y , X ) a~ phI Vys then it is easy to see that , -1 &lt; H &gt; Q ( tftt , / ( 311t , ut ) -b , -yt , j3 ce VTtE THEORY OF MATRICES .
29 and hence , from the foregoing values of ( Y 7)(T+f ) ' and ( T+i1)(1-T)- ' % we find II Q-l~t Q)(T+Q)1f2=( 1+2de c 2b , -2a3 2h , -1+2g , -2f , -2e 21 , 2k , -1-2j , -2i 2pe 20 o 2nz , -1-2m ; and -21_Q-l(T+Q)(t~~~ )(-2m , 2i , ~ 2e , 2a -29 -1,2j , 2f , 2b -2o.,.-2k , -1+2g , 2c -2p -21 , +2h , -1+2 , d which shows that the Matrix IT for the automorphic transformation of the funetion Xw'+yz'-zy'-wx ' is such that writing 11= A , B , C , D ) we haveTPI( P , L , -II , -D E , F , GI HI 0 , K G.--C I , J IK , L:-NI -J , F , B M , N. O. P -M , -I , E , A which is the theorem in question .
12 .
I remark in reference to the foregoing proof that writing T=(a , h , g , 1 ) h , b , f , i g , f , c , n ?
m , n , .
then the actual value of ( T+Q2)1 , =( a , h , -g ; , i-i ) ... h , b , f-i , m g , f+1 , c , n ~1+1 , mn , d ; is 1 A+d , H+n+v , G. , , LI+g H+n-Y ' , B+c , F-f +A , M-g+ G-m+pj , F-f-x , C+b6 , N+h+ L-I-0 , M-g-- , N+h , D+a where ( A , H , G , L ) H , B- , -F M G , F , C , N L , M , N , D is the matrix formed with the first minors of ( a , h , g , I ) ; h , 1 ; , f , rn g , f , c , n 1 , m , nd moreover A=ad-12 +nh-mg+1 , =6c-f2+nh -mg+I , p=6'n-mmf+dh-ml , o=tfg-ch+g1-a w =dg-nl +nf-cn , 7=hf-4g+ma-Ih 3 and A is the determinant ( a , h , , 1+1 ) h , b , f+1 , mCa A , f mn , dr , g , f , c 13 .
The expression for(T-f2)- &gt; is obtained from that of ( T +f &gt; )- ' by merely transposing the terms of the matrix , or , what is the same thing , by changing the signs of A , P5 'r .
And it would be easy by means of these developed values to verify the foregoing comparison of ( T- ?
)(T+Q2 &gt; 1 and ( T+2)(Tf Q)-1 .
Article Nos. 14 to 22.-Second Investigation .
14 .
I consider from a different , point of view the theory of a matrix 1=( a , b , c , d ) such that l-L=( p , 1 , -h , -d ) , e , f , ; fb A oS k lc-g , -c e , j , k , I -n , k , 6 M , n % 0 , p - , e , a or , as we may call it , a Hiermitian matrix .
Lemma .
The .
determinant V-aK b , G , d n , oj , p may be expressed , and that in two different ways , as a Pfaffian .
16 .
In fact multiplying the determinant into itself thus , V2=z a , 6 , bd tr .
d c , b , -a e.f , g , h h , g , -f , -e j , kS I 1 , kgJ.mg n , 0 , p p , o -n , -m we find ( d , c , -6 , -a ) , ( h y , -f -e ) , ( 1 , k , , ( p , o , -n , n ) V2_(a , b , c , d ) , =|Sill S121 813 ) S14 ( e , f , g , h ) 821 S229 p23 !
824 ( i , j , k , 1 ) , , , , I S31 , S3 S3 , 834 ( w7 , n , o , 9p|1 91 s+ , , S42 , S3 , 544 viz. we have si , =(a , 6 , c , d)(d , c , -b , -a ) , S12=( , 6 , c , dXh , g , -f , -e ) , &amp;c. : we see at once that S8=0 , 812+821-0 , &amp;c. , viz , the determinant in s is a skew determinant , that is , the square of a Pfaffian .
We have therefore V2= ( S12 8341+ S13 S42 + 814 823),9 or extracting the square root of each side , and determining the sign by a comparison of any single term , we have V=3s -S12 S.14 13 S42 S14 823 which is one of the required forms of V. 17 .
And in the same manner V2= tr .
a , b , c , d. in , n , P , p e , f , g , h , kj I tb,2 k , I -eS-f , -Sin , n , o , p -a which is equal to the determinant ( n , $-ea n j- , f , -1b ) , ( o k -gc ) ( _P 1 , -h , -d3 Itlll t125 t131 t14 ( a , e , m n)f , ... . .
t21 , t22 f231 t24 ( b6 , f , jn ) , , , , , , t31 t32 t33 t34 ( C , g , k , o ) , ... .
, t411 t42 ) t43 , t44 ( d , A , 1 , p ) I , , " ThEO2R O-E MARICES ' viz. t,1=(a , e , i , mn &gt; m , i , -e , -a ) , &amp;c. ; this is likewise a skew determinant , and we have V2=('t,2 t34+tI3 t24+t14 t23 ) , or extracting the square root of each side , and determining the sign by the comparisom of any single tern , we have Qt12 t4 + t'3 t24 +t14 t 3 , ,5 which is the .
other of the required form , s of V. 18 .
Considernow the matrix ( 'a , b , c , d ) e , f , 9g , h z , j , k , 1 qrn , n , 0 , p which is saich that e )f , g , h o , kr , -g , ---c i , j , k , 1 -n , -J , .
f , 6m n , 0 , r , e , a this , gives ( 1 , 0 , 0 , 0 )=( a , b , c , d )( p , 1 , -h , -d , 1 , 0 &gt; 0,0 e , f , g , h 0 , k , -g , -c 0 , 0- , 1 , 0 i , j , k , I -n-.n j , f , b 0 , 0 , 0 , 1 nr , , o0 p m , - , e , a ( p , o , -n , -rn ) , ( 1 , k , , ( -h , -gf , e ) ; ( -d , -c , 6 , a a , b , c , d ) , .
, .
( e , f , g7 , h ) , ... ( rn , n , o , p ) |,.,. .
, .
which is in fact ( 1 0 .
0,5 0 )(S14 S13 , -SI25 -'-SI1 )5:0 1 , 0 , 0 S24 , S233 S22 -821 0 , 0 , 1 , 0 834 , S33 , -S32 , -S31 a09 , 0 , 0 , 1 844 , S43 , -S42W -S41 and the two matrices will be equal , term by term , if only 1 14- " S23 , 0 S13..S12 S24 -S34 that is , if six conditionls are satisfied .
3c3 19 .
But we have also ( a matrix and its reciprocal being convertible ) ( 1 , 0 , 0 , 0( p , 1 , -h , -d )( a , b , c , d ) 0 , 1 , 00 o , J , -g , | e , f , g , h| 0 , 0 , 1 , 0 -n , j , f , b i , j , k , 1 0 , 0 , 0 , 1 -I , -t , e , a qn , n , o , p ( a , e , i , in ) , ( &amp; , f~j , n ) , ( c , g , k , o ) , ( d , h , l5 p)~ =(-n , 1j , f-h , -d ) , , , ,,9 ( o k , -g , c ) , , , , , ) , ( n j , f , b ) , , , , 1 , 7 ( -.m , -i , e , a ) , ... .
which is in fact 1 0 , 0 , 0 )=( t14 t24 t34 t44 ) 0 , 1 , 0 , 0 t13 , t23 t33 , t43 0 , 0 , 10 t125 -t22 -t32 ) t42 Oj 0O1 -till -41 t31 t41 and we obtain for the equality of the two matrices the six conditions 1=tI4=t23 , 0-t13= tl2= t24 t34 , equivalent to the former set of six conditions .
20 .
We obtain from either set of conditions , for the determinant the value Va , b , .
c , d X2 .
e , f,5 g5 i , j , k , I in , n , o , p 21 .
Write ( x , y , z , w)=( a , b , c , d )XX , Y , Z , W ) ; ( x ' , *y , z , w')=( C6 , 6 , c , d X , Y ' , Z ' , W X ) e , f , g , A e , f , g , h i , j , k , I i , , k , I m , n , o , ?
n , n , o , p then substituting for ( x , y , z , w)(x ' , y ' , z ' , w ' ) their values , we find xw'+yz'-zy'-wx'=-( t1 , , tM2 t13 , t14 3X , Y , Z , WXX ' , Y ' , Z ' , W ' ) t21 , t22 t23 , t24 t31 , t32 , t33 t34 t41 , t42 , t43f t44 =XW'+YZ ' -ZY'WX ' ; and similarly writing ( X , Y , Z , W)=( ?
p , 1 , -h , -d '(x , y , z , w ) , ( X ' , Y ' , Z ' , W')=( p , 1 , -h , -d )d , y ' , z ' , w ' ) , o , k , -9 , -c o , k , -g , -c -n , - , f , b -n , , , f , 6 M , - , e , a -rn , - , e , a we obtain with the s coefficients the equivalent result , XW'+YZ'-ZY'--WX'=xw'+yz'-zY-Wx ' .
We thus see conversely that the Hermitian matrix is in fact the matrix for the automorphic transformation of the function xw ' +yz ' zy ' wx ' .
22 .
Considering any two or more matrices for the automorphic transformation of such a function , the matrix compounded of these is a matrix for the automorphic transformation of the function-or , theorem , the matrix compounded of two or more Hermitian matrices is itself Hermitian .
Article No. 23.-Theorem on a Form of Matrices .
23 .
I take the opportunity of mentioning a theorem relating to the matrices which present themselves in the arithmetical theory of the composition of quadratic forms .
Writing ( X)= , a , a b+3 )and .
( X ) y '( y , , -c , b- ) |-ex , .
, b-/ 3 , c ||-Y , .
+* , -a , -(b-3 ) ) , .
, Y c , ( 6 -0+ ) , .
-(6+1 ) , -c , -y , .
7 -(6-1 ) , a , -1 , where D =ac-b2 , A-=cy_/ 32 ; and similarly , I I~ ~~~ ~~ r -'6'~-3 ( , x , a+ b13 ' ) and : .
( X')-'=D ' -7 ' Y , , + -a ' where D'=a'c ' b62 , A ' =y _[12 ; then or , what is the same thing , ( X"XX ' ) +(D-A)( U-A ' ) ( XiX ) ) is to a factor yres equal to the matrix unity ; viz. writing A=aca+2bP3+ cr+a'c ' +2143 ' +dy ' , the foregoing expression is =A( 1e The theorem is verified without difficulty by merely forming the expressions of the compound matrices ( XI(X ' ) and ( X''3X- ' ) .

