216 5. Further Determinant Theory
1 − 4 x 10 x
2
1 − 3 x 6 x
2
− 10 x
2
1 − 2 x 3 x
2
− 4 x
3
5 x
4
1 −xx
2
−x
3
x
4
−x
5
1 α 0 α 1 α 2 α 3 α 4 α 5
1 −xα 1 α 2 α 3 α 4 α 5 α 6
− 2 xx
2
α 2 α 3 α 4 α 5 α 6
3 x
2
−x
3
α 3 α 4 α 5 α 6
(5.6.13)
These determinants areB
66
s
,s=1, 2 ,3, as indicated at the corners of the
frames.B
66
1
is symmetric and is a bordered Hankelian. The dual identities
are found in the manner described in Theorem 5.16.
All the determinants described above are extracted from consecutive rows
and columns ofMorM
∗
. A few illustrations are sufficient to demonstrate
the existence of identities of a similar nature in which the determinants are
extracted from nonconsecutive rows and columns ofMorM
∗
.
In the first two examples, either the rows or the columns are nonconsec-
utive:
∣
∣
∣
∣
φ 0 φ 2
φ 1 φ 3
∣
∣
∣
∣
=−
∣ ∣ ∣ ∣ ∣ ∣
1 − 2 x
α 0 α 1 α 2
α 1 α 2 α 3
∣ ∣ ∣ ∣ ∣ ∣
, (5.6.14)
∣
∣
∣
∣
φ 1 φ 3
φ 2 φ 4
∣
∣
∣
∣
=
∣ ∣ ∣ ∣ ∣ ∣ ∣
1 − 2 x
1 α 0 α 1 α 2
−xα 1 α 2 α 3
x
2
α 2 α 3 α 4
∣ ∣ ∣ ∣ ∣ ∣ ∣
=
∣ ∣ ∣ ∣ ∣ ∣ ∣
1 − 3 x
1 −xx
2
−x
3
α 0 α 1 α 2 α 3
α 1 α 2 α 3 α 4
∣ ∣ ∣ ∣ ∣ ∣ ∣
.(5.6.15)
In the next example, both the rows and columns are nonconsecutive:
∣
∣
∣
∣
φ 0 φ 2
φ 2 φ 4
∣
∣
∣
∣
=−
∣ ∣ ∣ ∣ ∣ ∣ ∣
1 − 2 x
α 0 α 1 α 2
1 α 1 α 2 α 3
− 2 xα 2 α 3 α 4
∣ ∣ ∣ ∣ ∣ ∣ ∣
. (5.6.16)
The general form of these identities is not known and hence no theorem is
known which includes them all.
In view of the wealth of interrelations between the matricesMandM
∗
,
each can be described as the dual of the other.
Exercise.Verify these identities and their duals by elementary methods.
The above identities can be generalized by introducing a second variable
y. A few examples are sufficient to demonstrate their form.
φ 1 (x+y)=
∣
∣
∣
∣
1 −x
φ 0 (y) φ 1 (y)
∣
∣
∣
∣
=
∣
∣
∣
∣
1 −y
φ 0 (x) φ 1 (x)
∣
∣
∣
∣
, (5.6.17)
φ 1 (y)=
∣
∣
∣
∣
1 x
φ 0 (x+y) φ 1 (x+y)