48 3. Intermediate Determinant Theory
Proof.
Bij=(−1)
i+j
∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣
a 11 a 12 ··· a 1 ,j− 1 a 1 ,j+1 ··· a 1 n x 1
a 21 a 22 ··· a 2 ,j− 1 a 2 ,j+1 ··· a 2 n x 2
..........................................................
ai− 1 , 1 ai− 1 , 2 ··· ai− 1 ,j− 1 ai− 1 ,j+1 ··· ai− 1 ,n xi− 1
ai+1,i ai+1, 2 ··· ai+1,j− 1 ai+1,j+1 ··· ai+1,n xi+1
..........................................................
an 1 an 2 ··· an,j− 1 an,j+1 ··· ann xn
y 1 y 2 ··· yj− 1 yj+1 ··· yn z
∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ n
.
The expansion is obtained by applying arguments toBijsimilar to those
applied toBin Theorem 3.9. Since the second cofactor is zero whenr=i
ors=jthe double sum contains (n−1)
2
nonzero terms, as expected. It
remains to prove thatBij=Eij.
Transfer the last row ofBijto theith position, which introduces the sign
(−1)
n−i
and transfer the last column to thejth position, which introduces
the sign (−1)
n−j
. The result isEij, which completes the proof.
The Cauchy expansion of an arbitrary determinant focuses attention on
one arbitrarily chosen elementaijand its cofactor.
Theorem 3.11. The Cauchy expansion
A=aijAij+
n
∑
r=1
n
∑
s=1
aisarjAir,sj.
First Proof.The expansion is essentially the same as that given in Theorem
3.10. TransformEijback toAby replacingzbyaij,xrbyarjandysby
ais. The theorem appears after applying the relation
Air,js=−Air,sj. (3.7.2)
Second Proof.It follows from (3.2.3) that
n
∑
r=1
arjAir,sj=(1−δjs)Ais.
Multiply byaisand sum overs:
n
∑
r=1
n
∑
s=1
aisarjAir,sj=
n
∑
s=1
aisAis−
n
∑
s=1
δjsaisAis
=A−aijAij,
which is equivalent to the stated result.
Theorem 3.12. Ifys=1, 1 ≤s≤n, andz=0, then
n
∑
j=1
Bij=0, 1 ≤i≤n.