3.3 The Laplace Expansion 29
The expansion formula (3.3.4) follows.
Illustrations
1.Whenr= 2, the Laplace expansion formula can be proved as follows:
Changing the notation in the second line of (3.3.10),
A=
n
∑
p=1
n
∑
q=1
aipajqAij;pq,i<j.
This double sum containsn
2
terms, but thenterms in whichq=pare
zero by the definition of a second cofactor. Hence,
A=
∑
p<q
aipajqAij,pq+
∑
q<p
aipajqAij;pq.
In the second double sum, interchange the dummiespandqand refer
once again to the definition of a second cofactor:
A=
∑
p<q
∣
∣
∣
∣
aip aiq
ajp ajq
∣
∣
∣
∣
Aij;pq
=
∑
p<q
Nij;pqAij;pq,i<j,
which proves the Laplace expansion formula from rowsiandj. When
(n, i, j)=(4, 1 ,2), this formula becomes
A=N 12 , 12 A 12 , 12 +N 12 , 13 A 12 , 13 +N 12 , 14 A 12 , 14
+N 12 , 23 A 12 , 23 +N 12 , 24 A 12 , 24
+N 12 , 34 A 12 , 34.
2.Whenr= 3, begin with the formula
A=
n
∑
p=1
n
∑
q=1
n
∑
r=1
aipajqakrAijk,pqr, i<j<k,
which is obtained from the second line of (3.3.11) with a change in
notation. The triple sum containsn
3
terms, but those in whichp,q,
andrare not distinct are zero. Those which remain can be divided into
3! = 6 groups according to the relative magnitudes ofp,q, andr:
A=
[
∑
p<q<r
+
∑
p<r<q
+
∑
q<r<p
+
∑
q<p<r
+
∑
r<p<q
+
∑
r<q<p
]
aipajqakrAijk,pqr.