5.2 The Generalized Cusick Identities 185
From (5.2.4),
∂ai, 2 n
∂φ 2 n− 1
=ψi.
Also,
∂Hn
∂φ 2 n− 1
=Hn− 1.
Hence,
2 n− 1
∑
i=1
(−1)
i+1
Pf
(n)
i
ψi=Hn− 1 Kn. (5.2.24)
Similarly,
2 n− 1
∑
i=1
(−1)
i+1
Pf
(n)
i φi=HnKn−^1. (5.2.25)
The following three theorems express modified forms of|aij|nin terms of
the Hankelians.
LetBn(φ) denote the determinant which is obtained from|aij|nby
replacing the last row by the row
[
φ 1 φ 2 φ 3 ...φn
]
.
Theorem 5.6.
a.B 2 n− 1 (φ)=Hn− 1 HnK
2
n− 1 ,
b.B 2 n− 1 (ψ)=H
2
n− 1 Kn−^1 Kn,
c. B 2 n(φ)=−H
2
n
Kn− 1 Kn,
d.B 2 n(ψ)=−Hn− 1 HnK
2
n
.
Proof. ExpandingB 2 n− 1 (φ) by elements from the last row and their
cofactors and referring to (5.2.13), (5.2.14), and (5.2.25),
B 2 n− 1 (φ)=
2 n− 1
∑
j=1
φjA
(2n−1)
2 n− 1 ,j
=Pf
(n)
2 n− 1
2 n− 1
∑
i=1
(−1)
i+1
Pf
(n)
i φi
=Pfn− 1 HnKn− 1. (5.2.26)
Part (a) now follows from Theorem 5.1 and (b) is proved in a similar
manner.
ExpandingB 2 n(φ) with the aid of Theorem 3.9 on bordered determinants
(Section 3.7) and referring to (5.2.11) and (5.2.25),
B 2 n(φ)=−
2 n− 1
∑
i=1
2 n− 1
∑
j=1
ai, 2 nφjA
(2n−1)
ij