Acta Mathematica Scientia 2018,38B(6):1789–1804

http://actams.wipm.ac.cn

THE Cn WP-BAILEY CHAIN∗

Zhizheng ZHANG ( ܃)

Department of Mathematics, Luoyang Normal University, Luoyang 471934, China
E-mail : [email protected]

Junli HUANG ( ‘d|)

College of Mathematics and Information Science, Henan Normal University, Xinxiang 475001, China
E-mail : [email protected]

Abstract The purpose of this paper is to introduce the concept of Cn WP-Bailey pairs. The
Cn WP-Bailey transform is obtained by applying the Cn 6 φ5 summation formula. From this
result, the Cn WP-Bailey lemma is deduced by making use of the Cn q-Dougall summation
formula. Some applications are investigated. Finally, the case of elliptic Cn WP-Bailey pairs
is discussed.
Key words Cn WP-Bailey pair; Cn WP-Bailey chain; elliptic Cn WP-Bailey pair
2010 MR Subject Classification 33D67

1 Introduction
The classical Bailey transform and Bailey lemma play a very important role in the theory
and applications of the basic hypergeometric series [22, 43]. The Bailey transform was firstly
discovered by Bailey [7]. Slater [43] utilized it to obtain many Rogers-Ramanujan type identities.
Subsequently Andrews [2] established the iterative “Bailey chain” concept which led to a wide
range of applications. Building on prior work of Bressoud [14] and Singh [42], by introducing
a parameter k, Andrews [4] extended the Bailey pair to the new concept of WP-Bailey pair.
Andrews obtained two distinct ways to construct new WP-Bailey pairs from a given pair. In [5],
Andrews and Berkovich further investigated these two branches of the WP-Bailey tree, in the
process deriving many new transformations for basic hypergeometric series. Spiridonov [44, 45]
extended the Bailey chain to the setting of elliptic hypergeometric series. Warnaar [48] added
four more transformations which gave four more branches of the Bailey tree and two elliptic
WP-Bailey chains. Laughlin and Zimmer [25–28] obtained many results about WP-Bailey pairs.
Berkovich and Warnaar [9], Bressoud, Ismail and Stanton [16] discussed Bailey lemma viz the
change of base of Bailey pairs. For other related works, see [1, 15, 23, 24, 30, 31, 38, 39, 46].
Multiple basic hypergeometric series associated to the root systems An , Cn and Dn have
been investigated by many authors. The U (n + 1) (or An ) and Cn generalizations of the
Bailey lemma are due to Milne and Lilly [29, 34, 35]. In [8], Bartlett and Warnaar recently
∗ Received July 10, 2017; revised February 24, 2018. This research was supported by the National Natural

Science Foundation of China (11371184).

1790 ACTA MATHEMATICA SCIENTIA Vol.38 Ser.B

pointed out a correction in the statement of the Cn Bailey lemma in [29, 34, 35]. For the most
recent developments in the theory of multiple basic (and elliptic) hypergeometric series on root
systems, see Andrews [3], Andrews, Schilling and Warnaar [6], Bhatnagar [10, 18], Bhatnagar
and Milne [11], Bhatnagar and Schlosser [12], Coskun [17, 19, 20], Coskun and Gustafson [21],
Milne [32], Rains [36], Rosengren [37], Schlosser [40, 41], Warnaar [47, 49], Zhang [50], Zhang
and Liu [51], Zhang and Wu [52], etc.
The purpose of this paper is to introduce the concept of Cn WP-Bailey pairs. The Cn WP-
Bailey transform is obtained by applying the Cn 6 φ5 summation formula. From this result,
the Cn WP-Bailey lemma is deduced by making use of the Cn q-Dougall summation formula.
Some applications are investigated. Finally, the case of elliptic Cn WP-Bailey pairs is discussed.
Similar results for root systems An and Dn were very recently, after the completion of the present
paper, independently obtained by Bhatnagar and Schlosser [13].
We recall the standard basic hypergeometric notations from [22]. Let q be a complex
number such that |q| < 1. Define
∞
Y
(a)∞ := (a; q)∞ := (1 − aq k )
k=0

and
(a)∞
(a)n := (a; q)n := . (1.1)
(aq n )∞
Then

(a)n+k = (a)n (aq n )k , (1.2)

(a)n
(−q/a)k q ( 2)−nk .
k
(a)n−k = (1.3)
(q 1−n /a)k
We also adopt the following notation for multiple q-shifted factorials:

(a1 , a2 , · · · , am )N := (a1 , a2 , · · · , am ; q)N := (a1 ; q)N (a2 ; q)N · · · (am ; q)N ,

where N = Z ∪ {∞}. The basic hypergeometric series r φs is defined by

 
(a1 , a2 , · · · , ar )n h i1+s−r
X∞
a1 , a2 , · · · , ar
(−1)n q ( 2 )
n
r φs  ; q, z  = zn.
b1 , b2 , · · · , bs n=0
(q, b 1 , b 2 , · · · , b s )n

We abbreviate n-tuples such as (y1 , · · · , yn ) by y, and define |y| := y1 + · · · + yn . We will

also write y ≤ N to mean yi ≤ Ni for all 1 ≤ i ≤ n.
In [29], Lilly and Milne gave a Cn generalization of the Bailey lemma. They established
the following concept of Cn Bailey pairs.
Definition 1.1 Let x = (x1 , · · · , xn ) be an n-tuple of indeterminates and for N , y ∈ N n ,
let
n
Y 1
MN ,y (x) := yr −ys +1 x /x ) yr +ys +1 x x )
r,s=1
(q r s N r −y r (q r s Nr −yr

Yn
(qxr /xs )yr −ys (qxr xs )yr +ys
= .
r,s=1
(qxr /xs )Nr −ys (qxr xs )Nr +ys
No.6 Z.Z. Zhang & J.L. Huang: THE Cn WP-BAILEY CHAIN 1791

Then AN (x) and BN (x) form a Cn Bailey pair (AN (x), BN (x)) if for all N ≥ 0,
X
BN (x) = MN ,y (x)Ay (x). (1.4)
0≤y≤N

Lilly and Milne established the following inverse relation of (1.4).

Lemma 1.2 (see [29, Theorem 4.3]) Let x = (x1 , · · · , xn ) be an n-tuple of indeterminates
and, for N , y ∈ N n , we have
X
∗
AN (x) = MN ,y (x)By (x), (1.5)
0≤y≤N

where
|N |−|y| Y 1 − xr xs q yr +ys
∗ |N |−|y| ( )
MN ,y (x) : = (−1) q 2
1 − xr xs q Nr +Ns
1≤r<s≤n
n
Y 1
× yr −ys +1 x /x ) yr +Ns x x )
r,s=1
(q r s N r −y r (q r s Nr −yr

|N |−|y| Y 1 − xr xs q yr +ys
= (−1)|N |−|y| q ( 2 )
1 − xr xs q Nr +Ns
1≤r<s≤n
Yn
(qxr /xs )yr −ys (xr xs )yr +Ns
× . (1.6)
r,s=1
(qx r /xs )Nr −ys (xr xs )Nr +Ns

The most important ingredient of the theory is the Bailey lemma, which generates an
infinite sequence of Bailey pairs from an initial Bailey pair. But, Milne and Lilly’s Cn Bailey
lemma, first stated as [34, equation 2.8] and copied verbatim in [29] and [35] contains a minor
′
typographical error in the expression for BN (x). In the following this was corrected by Bartlett
and Warnaar [8].
Lemma 1.3 (Cn Bailey lemma, [8, Lemma 4.1]) If (AN (x), BN (x)) is a Cn Bailey pair,
then so is (A′N (x), BN
′
(x)) given by
 |N | Yn
′ q (αxi , βxi )Ni
AN (x) = AN (x) , (1.7)
αβ i=1
(qxi /α, qxi /β)Ni

X    |y| Yn
′ q q (αxi , βxi )yi
BN (x) = By (x)
αβ |N |−|y| αβ i=1
(qxi /α, qxi /β)Ni
0≤y≤N
Y n
(qxr xs )yr +ys Y (qxr /xs )yr −ys
× . (1.8)
(qxr xs )Nr +Ns r,s=1 (qxr /xs )Nr −ys
1≤r<s≤n

Our paper is organized as follows. In Section 2 we extend the above WP-Bailey pair to a Cn
WP-Bailey pair by applying Cn q-Dougall summation formula [35]. In Section 3 we establish the
Cn WP-Bailey chain. Section 4 begins with the unit Cn WP-Bailey pair and some applications
are given. In Section 5 the case of elliptic WP-Bailey pairs is discussed.

2 The Cn WP-Bailey Pairs

In this section, we introduce the concept of Cn WP-Bailey pair and its inverse relation. In
order to achieve our goal, we need to recall some results from Milne and Lilly [33, 35].
1792 ACTA MATHEMATICA SCIENTIA Vol.38 Ser.B

Lemma 2.1 (Cn q-Dougall summation formula [35, Theorem 6.7]) Let a, b, c, d, x1 , · · · ,
xn be indeterminates with abcd = q 1+|N | . Then
X n n
∆C (xq y ) Y (axr , bxr , cxr , dxr )yr Y (q −Ns xr /xs , xr xs )yr
q |y|
∆C (x) r=1 (qxr /a, qxr /b, qxr /c, qxr /d)yr r,s=1 (qxr /xs , q 1+Ns xr xs )yr
0≤y≤N
Y Yn
(qxr xs )Nr (qx2r )Nr
= (q/ab, q/ac, q/bc)|N | ,
(q 1+Ns xr xs )Nr r=1 (qxr /a, qxr /b, qxr /c, q 1+|N |−Nr /abcxr )Nr
1≤r<s≤n
(2.1)
Q
n Q
where ∆C (x) = (1 − x2r ) (xr − xs )(1 − xr xs ) and xq y = (x1 q y1 , · · · , xn q yn ).
r=1 1≤r<s≤n
It is clear when cd = q, there follows from (2.1).
Lemma 2.2 (Cn terminating 6 φ5 summation formula [35, Theorem 2.2]) Let a, b, x1 ,
· · · , xn be indeterminates. Then
X n n
∆C (xq y ) Y (axr , bxr )yr Y (q −Ns xr /xs , xr xs )yr
q |y| = δN ,0 , (2.2)
∆C (x) r=1 (qxr /a, qxr /b)yr r,s=1 (qxr /xs , q 1+Ns xr xs )yr
0≤y≤N

where ab = q |N | .
Let k, x1 , x2 , · · · , xn be indeterminates. Now we will define the WP analogues of the kernel
MN ,y (x) defining the Cn Bailey pair and its inverse relation. Set
n
Y n
Y 1
MN ,y (x|k) = (kxr )|N |+yr (k/xr )|N |−yr yr −ys +1 x /x ) yr +ys +1 x x )
r=1 r,s=1
(q r s N r −y r (q r s Nr −yr

Yn Yn
(qxr /xs )yr −ys (qxr xs )yr +ys
= (kxr )|N |+yr (k/xr )|N |−yr (2.3)
r=1 r,s=1
(qxr /xs )Nr −ys (qxr xs )Nr +ys

and
n
Y
|N |−|y| 1
∗
MN ,y (x|k) = (−1) |N |−|y|
q( 2 ) yr −ys +1 x /x ) y +Ns x x )
r,s=1
(q r s Nr −yr (q r r s Nr −yr

Y n
1 − xr xs q yr +ys Y (1 − kxr q |y|+yr )(1 − kq |y|−yr /xr )
×
1 − xr xs q Nr +Ns r=1 (kxr )|y|+Nr +1 (k/xr )|y|−Nr +1
1≤r<s≤n
Yn
|N |−|y| (qxr /xs )yr −ys (xr xs )Ns +yr
= (−1)|N |−|y| q ( 2 )
r,s=1
(qxr /xs )Nr −ys (xr xs )Nr +Ns
Y n
1 − xr xs q yr +ys Y (1 − kxr q |y|+yr )(1 − kq |y|−yr /xr )
× . (2.4)
1 − xr xs q Nr +Ns r=1 (kxr )|y|+Nr +1 (k/xr )|y|−Nr +1
1≤r<s≤n
∗
If the subscript vectors of MN ,y (x|k) and MN ,y (x|k) are ordered lexicographically, then
∗
MN ,y (x|k) and MN ,y (x|k) are infinite-dimensional lower-triangular matrices.
According to the method stated in [1], by applying the Cn q-Dougall summation formula,
we obtain the following result.
∗
Theorem 2.3 (Cn WP-Bailey transform) Let MN ,y (x|k) and MN ,y (x|k) be defined as
in (2.3) and (2.4). Then
X X
∗ ∗
MN ,y (x|k)My,M (x|k) = MN ,y (x|k)My,M (x|k)
M ≤y≤N M ≤y≤N
No.6 Z.Z. Zhang & J.L. Huang: THE Cn WP-BAILEY CHAIN 1793
n
Y
= δN ,M := δNr ,Mr , (2.5)
r=1

where δi,j denotes the classical Kronecker delta, that is δi,j = 1, if i = j, and 0 otherwise.
∗
Proof We start with the sum side of (2.5) where MN ,y (x|k) and MN ,y (x|k) are given by
(2.3) and (2.4). In the resulting multiple sum, shift the index of summation to 0 ≤ y ≤ N −M .
We then obtain
X
∗
MN ,y (x|k)My,M (x|k)
M ≤y≤N
X
∗
= MN ,y+M (x|k)My+M ,M (x|k)
0≤y≤N −M
X n
Y
|y|
= (−1) |y|
q( 2 ) (kxr )|N |+yr +Mr (k/xr )|N |−yr −Mr
0≤y≤N −M r=1
Yn
(qxr /xs )yr +Mr −ys −Ms (qxr xs )yr +Mr +ys +Ms
×
r,s=1
(qxr /xs )Nr −Ms −ys (qxr xs )Nr +Ms +ys
n
Y (qxr /xs )Mr −Ms (xr xs )ys +Ms +Mr
×
r,s=1
(qxr /xs )yr +Mr −Ms (xr xs )yr +Mr +ys +Ms
Y n
Y
1 − xr xs q Mr +Ms (1 − kxr q |M |+Mr )(1 − kq |M |−Mr /xr )
× y +M +y +M
.
1 − xr xs q r r s s r=1 (kxr )|M |+yr +Mr +1 (k/xr )|M |−yr −Mr +1
1≤r<s≤n

Applying the following relations

n
Y (kxr )|N |+yr +Mr (k/xr )|N |−yr −Mr
r=1
(kxr )|M |+yr +Mr +1 (k/xr )|M |−yr −Mr +1
n
Y
(|M |−|N |+1)|y| (kxr )|N |+Mr (k/xr )|N |−Mr
=q
r=1
(kxr )|M |+Mr +1 (k/xr )|M |−Mr +1

Yn
(kxr q |N |+Mr )yr (q −|M |+Mr xr /k)yr
× , (2.6)
r=1
(kxr q |N |+Mr +1 )yr (q 1−|N |+Mr xr /k)yr

Yn Yn
(qxr /xs )yr +Mr −ys −Ms 1
= 1+M −M +y −y
r,s=1
(qxr /xs )Nr −ys −Ms r,s=1
(q r s r s x /x )
r s Nr −Mr −yr

Yn
|y| (qxr /xs )Mr −Ms
= (−1)|y| q (|N |−|M |)|y|−( 2 )
r,s=1
(qxr /xs )Nr −Ms
Y n
xr q yr +Mr − xs q ys +Ms Y Mr −Ns
× (q xr /xs )yr (2.7)
xr q Mr − xs q Ms r,s=1
1≤r<s≤n

(see [33, (3.32)]),

n
Y (qxr /xs )Mr −Ms (xr xs )ys +Ms +Mr
r,s=1
(qxr /xs )yr +Mr −Ms (xr xs )yr +Mr +ys +Ms
n
Y 1
= (2.8)
r,s=1
(q Mr −Ms +1 xr /xs )yr (xr xs q Mr +Ms +ys )yr
1794 ACTA MATHEMATICA SCIENTIA Vol.38 Ser.B

and
Yn Yn Y
(qxr xs )yr +Mr +ys +Ms 1 − x2r q 2(yr +Mr ) (1 − xr xs q yr +ys +Mr +Ms )2
=
r,s=1
(qxr xs )Nr +Ms +ys r=1
1 − x2r q 2Mr (1 − xr xs q Mr +Ms )2
1≤r<s≤n
n
Y Mr +Ms
(xr xs q )ys (xr xs q Mr +Ms +ys )yr
× , (2.9)
r,s=1
M +M +1
(xr xs q r s )Nr −Mr (xr xs q Nr +Ms +1 )ys
we obtain
n
Y Yn
(qxr /xs )Mr −Ms (qxr xs )Mr +Ms
LHS of (2.5) = (kxr )|N |+Mr (k/xr )|N |−Mr
r=1 r,s=1
(qxr /xs )Nr −Ms (qxr xs )Nr +Ms
Yn
(1 − kxr q |M |+Mr )(1 − kq |M |−Mr /xr )
×
r=1
(kxr )|M |+Mr +1 (k/xr )|M |−Mr +1
X Yn
|y| 1 − x2r q 2(Mr +yr )
× q
r=1
1 − x2r q 2Mr
0≤y≤N −M
Yn
(kxr q |N |+Mr )yr (q −|M |+Mr xr /k)yr
×
r=1
(q 1−|N |+Mr xr /k)yr (kxr q |M |+Mr +1 )yr
Yn
(q Mr −Nr xr /xs )yr (xr xs q Mr +Ms )yr
×
r,s=1
(q r −Ms +1 xr /xs )yr (xr xs q Mr +Ns +1 )yr
M

Y xr q Mr +yr − xs q Ms +ys Y 1 − xr xs q Mr +Ms +yr +ys

×
xr q Mr − xs q Ms 1 − xr xs q Mr +Ms
1≤r<s≤n 1≤r<s≤n
n
Y Yn
(qxr /xs )Mr −Ms (qxr xs )Mr +Ms
= (kxr )|N |+Mr (k/xr )|N |−Mr
r=1 r,s=1
(qxr /xs )Nr −Ms (qxr xs )Nr +Ms
Yn
(1 − kxr q |M |+Mr )(1 − kq |M |−Mr /xr )
×
r=1
(kxr )|M |+Mr +1 (k/xr )|M |−Mr +1
X n
∆C (xq y+M ) Y (kxr q |N |+Mr , q −|M |+Mr xr /k)yr
× q |y|
∆C (xq M ) r=1 (q 1−|N |+Mr xr /k, kxr q |M |+Mr +1 )yr
0≤y≤N −M
n
Y Mr −Nr
(q xr /xs , xr xs q Mr +Ms )yr
× .
r,s=1
(q Mr −Ms +1 xr /xs , xr xs q Mr +Ns +1 )yr

By the Cn terminating 6 φ5 summation formula (2.2) with

(Nr , xr , a, b) → (Nr − Mr , xr q Mr , kq |N | , q −|M | /k),
we have
n
Y Yn
(qxr /xs )Mr −Ms (qxr xs )Mr +Ms
LHS of (2.5) = δN ,M (kxr )|N |+Mr (k/xr )|N |−Mr
r=1 r,s=1
(qxr /xs )Nr −Ms (qxr xs )Nr +Ms
n
Y 1
×
r=1
(kxr )|M |+Mr
(k/xr )|M |−Mr

= δN ,M = RHS of (2.5).
This completes the proof. 
An extension of Theorem 2.3 to general sequences of parameters was given by Schlosser in
[40, Theorem 4.1].
No.6 Z.Z. Zhang & J.L. Huang: THE Cn WP-BAILEY CHAIN 1795

Equations (2.3) and (2.4) motivate the definition of a Cn WP-Bailey pair.

Definition 2.4 Let x = (x1 , · · · , xn ) be an n-tuple of indeterminates and define MN ,y (x|k)
as in (2.3). Then AN (x; k) and BN (x; k) form a Cn WP-Bailey pair (AN (x; k), BN (x; k)) if
for all N ≥ 0
X
BN (x; k) = MN ,y (x|k)Ay (x; k). (2.10)
0≤y≤N

We have the following result by applying Theorem 2.3.

Lemma 2.5 (Cn WP-Bailey pair inversion) If AN (x; k) and BN (x; k) form a Cn WP-
Bailey pair (AN (x; k), BN (x; k)), we have
X
∗
AN (x; k) = MN ,y (x|k)By (x; k). (2.11)
0≤y≤N

3 The Cn WP-Bailey Chain

As an application of the Cn WP-Bailey pair, we next establish the concept of the Cn
WP-Bailey chain.
Theorem 3.1 (The Cn generalization of WP-Bailey lemma) Let A(x; k) = {AN (x; k)}
and B(x; k) = {BN (x; k)} form a Cn WP-Bailey Pair. If we define
 q |N | Yn
(axr , bxr )Nr
A′N (x; k) = AN (x; kab/q) (3.1)
ab r=1
(qxr /a, qxr /b)Nr

and
′
X (q/ab)|N |−|y|  q |y|
BN (x; k) = (ka, kb)|N | By (x; kab/q)
(ka, kb)|y| ab
0≤y≤N
Yn n
(1 − kabxr q |y|+yr −1 )(axr , bxr )yr Y (qxr /xs )yr −ys
×
r=1
(1 − kabx
q )(qxr /a, qxr /b)Nr
r
r,s=1
(qxr /xs )Nr −ys
Y n
(qxr xs )yr +ys Y (kxr )|N |+yr (k/xr )|N |−Nr
× , (3.2)
(qxr xs )Nr +Ns r=1 (kabxr )Nr +|y| (kab/qxr )|y|−yr
1≤r<s≤n

then A′ (x; k) = {A′N (x; k)} and B ′ (x; k) = {BN

′
(x; k)} also forms a Cn WP-Bailey pair.
Proof Applying (2.3), (3.1) and Lemma 2.5, changing the order of summation and making
some simplification by applying (2.7), (2.9) and (1.2), (1.3), it follows that
X
MN ,y (x|k)A′y (x; k)
0≤y≤N
X  q |m| Yn
(qxr /xs )mr −ms (qxr xs )mr +ms
= Bm (x; kαβ/q)
ab r,s=1
(qxr /xs )Nr −ms (qxr xs )Nr −ms
0≤m≤N
n
Y (axr , bxr )mr (kxr )|N |+mr (k/xr )|N |−mr
×
r=1
(qxr /a, qxr /b)mr (kabxr /q)|m|+mr (kab/qxr )|m|−mr
X n
∆C (xq m+y ) |y| Y (q mr −Ns xr /xs , xr xs q mr +ms )ys
× q
∆C (xq m ) r,s=1
(xr xs q 1+ms +Nr , q 1+mr −ms xr /xs )yr
0≤y≤N −m
Yn
(kxr q |N |+mr , axr q mr , bxr q mr , q 1−|m|+mr xr /kab)yr
× 1−|N |+mr x /k, q 1+mr x /a, q 1+mr x /b, q |m|+mr kabx )
.
r=1
(q r r r r yr
1796 ACTA MATHEMATICA SCIENTIA Vol.38 Ser.B

By the Cn q-Dougall summation formula (2.1) with

(Nr , xr , a, b, c, d) 7→ (Nr − mr , xr q mr , kq |N | , q 1−|m| /kab, b, a),

we obtain
X
Mx (N ; y|k; Cn )A′y (x; k)
0≤y≤N
X  q |m| Yn
(qxr /xs )mr −ms (qxr xs )mr +ms
= Bm (x; kab/q)
ab r,s=1
(qxr /xs )Nr −ms (qxr xs )Nr −ms
0≤m≤N
n h
Y i
(axr , bxr )mr (kxr )|N |+mr (k/xr )|N |−mr
×
r=1
(qxr /a, qxr /b)mr (kabxr /q)|m|+mr (kab/qxr )|m|−mr
|m|−|N |
×(q ab)|N |−|m| (q 1−|N | /kb)|N |−|m|(q |m| ka)|N |−|m|
Y (xr xs q 1+mr +ms )Nr −mr
×
(xr xs q 1+mr +Ns )Nr −mr
1≤r<s≤n
Yn
(x2r q 1+2mr )Nr −mr
×
r=1
(q 1+mr −|N | xr /k, q |m|+mr kxr ab, q 1+mr xr /b, q −Nr a/xr )Nr −mr
X (q/ab)|N |−|m|  q |m|
= (ka, kb)|N | Bm (x; kab/q)
(ka, kb)|m| ab
0≤m≤N
Yn n
(1 − kabxr q |m|+mr −1 )(axr , bxr )mr Y (qxr /xs )mr −ms
×
r=1
(1 − kabx
q )(qxr /a, qxr /b)Nr
r
r,s=1
(qxr /xs )Nr −ms
Y n
(qxr xs )mr +ms Y (kxr )|N |+mr (k/xr )|N |−Nr
×
(qxr xs )Nr +Ns r=1 (kabxr )Nr +|m| (kab/qxr )|m|−mr
1≤r<s≤n
′
= BN (x; k).

This completes the proof. 

Just as ordinary Bailey pairs [4, 46], the Cn WP-Bailey pairs are expected to have wide
applications in number theory, analysis, physics. From such Cn WP-Bailey pairs (AN (x; k),
BN (x; k)) in Theorem 3.1 it is possible to construct new Cn WP-Bailey pairs. By the iteration
of Cn WP-Bailey pairs, we can obtain the Cn WP-Bailey chain:

(AN (x; k), BN (x; k)) → (A′N (x; k), BN

′
(x; k)) → (A′′N (x; k), BN
′′
(x; k)) → · · · .

Hence, given an initial Cn WP-Bailey pair, one immediately finds an infinite sequence of Cn
WP-Bailey pairs. This sequence forms a Cn WP-Bailey chain.

4 Some Applications
Choosing

BN (x; k) = δ|N |,0 , (4.1)

using (2.11) we have

Y Yn
|N | 1 − xr xs 1
AN (x; k) = (−1)|N | q ( 2 )
N +N
1 − xr xs q r s
r,s=1
(qxr /xs )Nr (xr xs q Ns )Nr
1≤r<s≤n
No.6 Z.Z. Zhang & J.L. Huang: THE Cn WP-BAILEY CHAIN 1797
n
Y 1
×
r=1
(qkxr )Nr (qk/xr )−Nr
Y Yn
|N | 1 − xr xs 1
= (−1)|N | q ( 2 )
1 − xr xs q Nr +Ns r,s=1 (qxr /xs )Nr (xr xs q Ns )Nr
1≤r<s≤n
n
Y 1−Nr
(q k/xr )Nr
× . (4.2)
r=1
(qkxr )Nr

This is the unit Cn WP-Bailey pair. From this we will prove a Cn Bailey transformation.
Inserting the unit Cn WP-Bailey pair (4.1)–(4.2) into the Cn WP-Bailey lemma (3.1)–(3.2),
we get the Cn WP-Bailey pair
|N |
 q |N | Yn
(axr , bxr )Nr Y 1 − xr xs
A′N (x; k) = (−1)|N | q ( 2 )
ab r=1
(qxr /a, qxr /b)Nr 1 − xr xs q Nr +Ns
1≤r<s≤n
n
Y Yn
1 (q −Nr kab/xr )Nr
× N
, (4.3)
r,s=1
(qxr /xs )N r (xr xs q s )Nr r=1
(kabxr )Nr

n
Y Yn
′ 1 1
BN (x; k) = (ka, kb, q/ab)|N |
r=1
(qxr /a, qxr /b)Nr r,s=1 (qxr /xs )Nr
Y Yn
1 (kxr , k/xr )|N |
× . (4.4)
(qxr xs )Ns (xr xs q 1+Ns )Nr r=1 (kabxr , kq |N |−Nr /xr )Nr
1≤r<s≤n

Substituting the Cn WP-Bailey pair (4.3)–(4.4) into the Cn WP-Bailey lemma (3.1)–(3.2), after
some simple manipulations, we obtain the Cn WP-Bailey pair A′′N and BN ′′

|N |
 q 2 |N | Yn
(axr , bxr , cxr , dxr )Nr
A′′N (x; k) = (−1) |N |
q( 2 )
abcd r=1
(qxr /a, qxr /b, qxr /c, qxr /d)Nr
Y Yn
1 − xr xs 1
× N +N
1 − xr xs q r s
r,s=1
(qxr /xs )Nr (xr xs q Ns )Nr
1≤r<s≤n
Yn −1−Nr
(q kabcd/xr )Nr
× (4.5)
r=1
(kabcdxr /q)Nr

and
n
Y Yn
′′ (kxr , k/xr )|N | 1
BN (x; k) = (kc, kd, q/cd)|N | |N |−Nr /x )
r=1
(qxr /c, qxr /d, kcdx r , kq r Nr r,s=1 (qxr /x s )Nr

Y X (kacd/q, kbcd/q, q/ab)|y| |y|

× (qxr xs )Nr +Ns q
1≤r<s≤n 0≤y≤N
(kc, kd, q −|N | cd)|y|
Yn Y n
1−q |y|+yr −1
kcdxr xr q yr − xs q ys Y (kcdxr /q, kcd/qxr )|y|
×
r=1
1 − kcdxr /q xr − xs r=1
(q Nr kcdxr , kcd/qxr )|y|
1≤r<s≤n
n
Y |N | Yn
(cxr , dxr , q kxr )yr (q −Ns xr /xs )yr
× . (4.6)
r=1
qxr /a, qxr /b, kabcdxr /q)yr r,s=1 (qxr /xs )yr

Upon substituting the WP-Bailey pair A′′N and BN

′′
into (2.10), and putting k = q 2 /abcde
|N |+2
and f = q /abcde, we obtain
1798 ACTA MATHEMATICA SCIENTIA Vol.38 Ser.B

Theorem 4.1 (Cn 10 φ9 transformation, [12, Theorem 2.1]) Let a, b, c, d, e, f , x1 , · · · ,

xn be indeterminates. Then
X ∆C (xq y ) Yn
(axr , bxr , cxr , dxr , exr , f xr )yr
q |y|
∆C (x) r=1
(qxr /a, qxr /b, qxr /c, qxr /d, qxr /e, qxr /f )yr
0≤y≤N
Yn
(q −Ns xr /xs , xr xs )yr
×
r,s=1
(q 1+Ns xr xs , qxr /xs )yr
Y Yn
1
= (q 2 /abde, q 2 /abce, q/cd)|N | (qxr xs )Nr
(qxr xs )Nr +Ns r,s=1
1≤r<s≤n
n
Y 1
×
r=1
(q 2 xr /abe, f q −Nr /xr , qxr /c, qxr /d)Nr
X Yn
∆(xq y ) |y| (q/ae, q/be, q/ab)|y| abe − q |y|+yr +1 xr
× q 2 2 2
∆(x) (q /abde, q /abce, q /abef )|y| r=1 abe − qxr
0≤y≤N
n
Y n
−Ns
(q xr /xs )yr Y (cxr , dxr , f xr )yr (qxr /abe)|y|
× ,
r,s=1
(qxr /xs )yr r=1 (qxr /a, qxr /b, qxr /e)yr (q Nr +2 xr /abe)|y|
Q
where abcdef = q |N |+2 and ∆(x) = (xr − xs ) is the Vandermonde product.
1≤r<s≤n
In the case when n = 1, by simplification this theorem reduces to Bailey’s transformation
formula for a terminating 10 φ9 series [22, (2.9.1)]:
 
x2 , qx, −qx, ax, bx, cx, dx, ex, f x, q−N
 
10 φ9  qx qx qx qx qx qx 1+N 2 q, q 
x, −x, , , , , , ,q x
a b c d e f
(q 2 /abde, q 2 /abce, q/cd, qx2 )N
= 2
(q x/abe, q 2 /abcdex, qx/c, qx/d)N
 r r 
qx qx qx q q q −N
 abe , q abe , −q abe , cx, dx, be , ae , ab , f x, q 
 
× 10 φ9  r r q, q , (4.7)
 qx qx q 2
q 2
qx qx qx q 2
q N +2 
,− , , , , , , ,
abe abe abce abde a b e abef abe
where abcdef = q N +2 .
Taking e → ∞ and f → 0 in Theorem 4.1, we have
Corollary 4.2 (Cn 8 φ7 transformation, [35, Theorem 6.6]) Let a, b, c, d, x1 , · · · , xn be
indeterminates. Then
X ∆C (xq y )  q |N |+2 |y| Y
n
(q −Ns xr /xs , xr xs )y r

∆C (x) abcd r,s=1

(q 1+Ns xr xs , qxr /xs )yr
0≤y≤N
Yn
(axr , bxr , cxr , dxr )yr
×
r=1
(qxr /a, qxr /b, qxr /c, qxr /d)yr
Y Yn Yn
1 1
= (q/cd)|N | (qxr xs )Nr
(qxr xs )Nr +Ns r,s=1 r=1
(qxr /c, qxr /d)Nr
1≤r<s≤n
X n n
∆(xq y ) |y| (q/ab)|y| Y (q −Ns xr /xs )yr Y (cxr , dxr )yr
× q .
0≤y≤N
∆(x) (q −|N | cd)|y| r,s=1 (qxr /xs )yr r=1 (qxr /a, qxr /b)yr
No.6 Z.Z. Zhang & J.L. Huang: THE Cn WP-BAILEY CHAIN 1799

In the case when n = 1, after a bit simplification the last identity reduces to the Watson’s
transformation formula for a terminating very-well-poised 8 φ7 series [22, (2.5.1)]:
 
x2 , xq, −xq, ax, bx, cx, dx, q −N q N +2
 
8 φ7  qx qx qx qx 1+N 2 q, abcd 
x, −x, , , , ,q x
a b c d
 q 
2 cx, dx, , q −N
(q/cd, qx )N  ab 
= 4 φ3  q, q  . (4.8)
(qx/c, qx/d)N qx qx −N
, , q cd
a b

5 The Elliptic Case

In [44], Spiridonov had extended the concept of WP-Bailey chains to elliptic hypergeometric
series and even gave an integral analogue (with contains the discrete case involving summations
in a special case) of the WP-Bailey lemma. Since elliptic hypergeometric series contain well-
poised basic hypergeometric series it is just natural to consider elliptic extensions of results
valid for well-poised basic hypergeometric series. This is also true for Cn series. In this section
we extend our results to the elliptic case. Our elliptic extensions involve a fixed parameter p
such that |p| < 1.
We denote by θ(z; p) the modified Jacobi theta function
∞
Y
θ(z; p) = (1 − zpj )(1 − pj+1 /z)
j=0

and define the elliptic q-shifted factorial by the product

n−1
Y
(a; q, p)0 = 1, (a; q, p)n = θ(aq j ; p)
j=0

and
1
(a; q, p)−n = ,
(aq −n ; q, p)n
where n = 1, 2, · · · .
As usual we employ the condensed notation
(a1 , · · · , ak ; q, p)n = (a1 ; q, p)n · · · (ak ; q, p)n .
The very-well-poised elliptic hypergeometric series are defined by
∞
X θ(a1 q 2k ; p) (a1 , a6 , · · · , ar+1 ; q, p)k q k
r+1 Vr (a1 ; a6 , · · · , ar+1 ; q, p) = .
θ(a1 ; p) (q, a1 q/a6 , · · · , a1 q/ar+1 ; q, p)k
k=0

In [37], Rosengren established the elliptic Cn Jackson summation [37, Theorem 7.1]:
X Yn Y
θ(x2r q 2yr ; p) q ys θ(q yr −ys xr /xs ; p)θ(xr xs q yr +ys ; p)
q |y|
r=1
θ(x2r ; p) θ(xr /xs ; p)θ(xr xs ; p)
0≤y≤N 1≤r<s≤n
n
Y n
(q−Ns
xr /xs , xr xs ; q, p)yr Y (bxr , cxr , dxr , exr ; q, p)yr
× 1+Ns x x ; q, p)
r,s=1
(qx r /xs , q r s yr r=1 (qxr /b, qxr /c, qxr /d, qxr /e; q, p)yr
n
Y Y 1
= (q/bc, q/d, q/cd; q, p)|N | (qxr xs ; q, p)Nr
r,s=1
(qxr xs ; q, p)Nr +Ns
1≤r<s≤n
1800 ACTA MATHEMATICA SCIENTIA Vol.38 Ser.B

n
Y 1
× , (5.1)
r=1
(qxr /b, qxr /c, qxr /d, q −Nr e/xr ; q, p)Nr

where q 1+|N | = bcde. We now specialize de = q , thus q |N | = bc, and obtain

X Yn Y
|y| θ(x2r q 2yr ; p) q ys θ(q yr −ys xr /xs ; p)θ(xr xs q yr +ys ; p)
q
r=1
θ(x2r ; p) θ(xr /xs ; p)θ(xr xs ; p)
0≤y≤N 1≤r<s≤n
n
Y n
(q xr /xs , xr xs ; q, p)yr Y (bxr , cxr ; q, p)yr
−Ns
× = δN ,0 . (5.2)
r,s=1
(qxr /xs , q 1+Ns xr xs ; q, p)yr r=1 (qxr /b, qxr /c; q, p)yr

Set
n
Y
MN ,y (x|k; q, p) = (kxr ; q, p)|N |+yr (k/xr ; q, p)|N |−yr
r=1
Yn
(qxr /xs ; q, p)yr −ys (qxr xs ; q, p)yr +ys
× (5.3)
r,s=1
(qx r /xs ; q, p)Nr −ys (qxr xs ; q, p)Nr +ys

and
Yn
∗ |N |−|y| ( |N |−|y|
) (qxr /xs ; q, p)yr −ys (xr xs ; q, p)Ns +yr
MN ,y (x|k; q, p) = (−1) q 2

r,s=1
(qx r /xs ; q, p)Nr −ys (xr xs ; q, p)Nr +Ns

Y n
θ(xr xs q yr +ys ; p) Y θ(kxr q |y|+yr ; p)θ(kq |y|−yr /xr ; p)
× .
θ(xr xs q Nr +Ns ; p) r=1 (kxr ; q, p)|y|+Nr +1 (k/xr ; q, p)|y|−Nr +1
1≤r<s≤n
(5.4)
Similar to the proof of Theorem 2.3, by applying (5.2) we obtain
X X
∗ ∗
MN ,y (x|k; q, p)My,M (x|k; q, p) = MN ,y (x|k; q, p)My,M (x|k; q, p)
M ≤y≤N M ≤y≤N

= δN ,M . (5.5)

Equations (5.3) and (5.4) motivates the definition of the elliptic Cn WP-Bailey pair.
Definition 5.1 (The elliptic Cn WP-Bailey pair) AN (x|k; q, p) and BN (x|k; q, p) form
an elliptic WP-Bailey pair if
X
BN (x|k; q, p) = MN ,y (x|k; q, p)Ay (x|k; q, p), (5.6)
0≤y≤N

Definition 5.1 and (5.1) immediately yield.

Lemma 5.2 (The elliptic Cn WP-Bailey pair inversion) If AN (x|k; q, p) and BN (x|k; q, p)
form an elliptic WP-Bailey pair, then we have
X
∗
AN (x|k; q, p) = MN ,y (x|k; q, p)By (x|k; q, p). (5.7)
0≤y≤N

Similar to the proof procedure of Theorem 3.1, by applying (5.1), we obtain

Theorem 5.3 (The elliptic Cn WP-Bailey lemma) Let A(x|k; q, p) = {AN (x|k; q, p)}
and B(x|k; q, p) = {BN (x|k; q, p)} form an elliptic Cn WP-Bailey Pair. If we define
 q |N | Yn
(axr , bxr ; q, p)Nr
A′N (x|k; q, p) = AN (x|kab/q; q, p) ,
ab r=1
(qxr /a, qxr /b; q, p)Nr
No.6 Z.Z. Zhang & J.L. Huang: THE Cn WP-BAILEY CHAIN 1801

′
X (q/ab; q, p)|N |−|y|  q |y|
BN (x|k; q, p) = (ka, kb; q, p)|N | By (x|kab/q; q, p)
(ka, kb; q, p)|y| ab
0≤y≤N
Yn n
θ(kabxr q |y|+yr −1 ; p)(axr , bxr ; q, p)yr Y (qxr /xs ; q, p)yr −ys
×
r=1
θ( kabx
q ; p)(qxr /a, qxr /b; q, p)Nr
r
r,s=1
(qxr /xs ; q, p)Nr −ys
Y (qxr xs ; q, p)yr +ys
×
(qxr xs ; q, p)Nr +Ns
1≤r<s≤n
Yn
(kxr ; q, p)|N |+yr (k/xr ; q, p)|N |
× ,
r=1
(kabxr ; q, p)Nr +|y| (kab/qxr ; q, p)|y|−yr (kq |N |−Nr /xr ; q, p)Nr
then A′ (x|k; q, p) = {A′N (x|k; q, p)} and B ′ (x|k; q, p) = {BN
′
(x|k; q, p)} also form an elliptic
Cn WP-Bailey pairs.
Choosing
BN (x|k; q, p) = δ|N |,0 , (5.8)
we have by (5.7)
n
Y
|N | (xr xs ; q, p)Ns
AN (x|k; q, p) = (−1)|N | q ( 2 )
r,s=1
(qxr /xs ; q, p)Nr (xr xs ; q, p)Nr +Ns
Y Yn
θ(xr xs ; p) 1
× . (5.9)
θ(xr xs q Nr +Ns ; p) r=1 (qkxr ; q, p)Nr (qk/xr ; q, p)−Nr
1≤r<s≤n

Substituting such a Bailey pair into the identity of Theorem 5.3, we get the elliptic Cn WP-
Bailey pair
|N |
 q |N | Yn
(axr , bxr ; q, p)Nr
A′N (x|k; q, p) = (−1)|N | q ( 2 )
ab r=1
(qxr /a, qxr /b; q, p)Nr

Y Yn
θ(xr xs ; p) 1
×
θ(xr xs q Nr +Ns ; p) r,s=1 (qxr /xs ; q, p)Nr (xr xs q Ns ; q, p)Nr
1≤r<s≤n
Yn −Nr
(q kab/xr ; q, p)Nr
× , (5.10)
r=1
(kabxr ; q, p)Nr

n
Y
′ 1
BN (x|k; q, p) = (ka, kb, q/ab; q, p)|N |
r=1
(qxr /a, qxr /b; q, p)Nr
n
Y Y
1 1
×
r,s=1
(qxr /xs ; q, p)Nr (qxr xs ; q, p)Ns (xr xs q 1+Ns ; q, p)Nr
1≤r<s≤n
n
Y (kxr , k/xr ; q, p)|N |
× |N |−Nr /x ; q, p)
. (5.11)
r=1
(kabxr , kq r Nr

Substituting the elliptic WP-Bailey pair into Theorem 5.3, after some simple manipulations,
we obtain the following elliptic WP-Bailey pair A′′N and BN ′′
,
|N |
 q 2  |N | Yn
(axr , bxr , cxr , dxr ; q, p)Nr
A′′N (x|k; q, p) = (−1)|N | q ( 2 )
abcd r=1
(qxr /a, qxr /b, qxr /c, qxr /d; q, p)Nr
Y Yn
θ(xr xs ; p) 1
× N +N
θ(xr xs q r s ; p) r,s=1 (qxr /xs ; q, p)Nr (xr xs q Ns ; q, p)Nr
1≤r<s≤n
1802 ACTA MATHEMATICA SCIENTIA Vol.38 Ser.B

Yn
(q −1−Nr kabcd/xr ; q, p)Nr
× , (5.12)
r=1
(kabcdxr /q; q, p)Nr

and
n
Y
′′ (kxr , k/xr ; q, p)|N |
BN (x|k; q, p) = (kc, kd, q/cd; q, p)|N |
r=1
(qxr /c, qxr /d, kcdxr , kq |N |−Nr /xr ; q, p)Nr
n
Y Y
1
× (qxr xs ; q, p)Nr +Ns
r,s=1
(qxr /xs ; q, p)Nr
1≤r<s≤n
X n
(kacd/q, kbcd/q, q/ab; q, p)|y| |y| Y θ(q |y|+yr −1 kcdxr ; p)
× q
0≤y≤N
(kc, kd, q −|N | cd; q, p)|y| r=1
θ(kcdxr /q; p)
Y n
q ys θ(q yr −ys xr /xs ; p) Y (kcdxr /q, kcd/qxr ; q, p)|y|
×
θ(xr /xs ; p) r=1
(q Nr kcdxr , kcd/qxr ; q, p)|y|
1≤r<s≤n
Yn Yn
(cxr , dxr , q |N | kxr ; q, p)yr (q −Ns xr /xs ; q, p)yr
× . (5.13)
r=1
(qxr /a, qxr /b, kabcdxr /q; q, p)yr r,s=1 (qxr /xs ; q, p)yr

Upon substituting the elliptic Cn WP-Bailey pair A′′N and BN

′′
into (5.6), and putting
2 |N |+2
k = q /abcde and f = q /abcde we obtain
Theorem 5.4 (An elliptic 10 φ9 transformation, [37, Corollary 8.3]) Let a, b, c, d, e, f ,
x1 , · · · , xn be indeterminates. Then
X Yn Y
θ(x2r q 2yr ; p) q ys θ(q yr −ys xr /xs ; p)θ(xr xs q yr +ys ; p)
q |y|
r=1
θ(x2r ; p) θ(xr /xs ; p)θ(xr xs ; p)
0≤y≤N 1≤r<s≤n
Yn n
(q xr /xs , xr xs ; q, p)yr Y
−Ns
(axr , bxr , cxr , dxr , exr , f xr ; q, p)yr
× 1+Ns x x , qx /x ; q, p)
r,s=1
(q r s r s y r r=1 (qxr /a, qxr /b, qxr /c, qxr /d, qxr /e, qxr /f ; q, p)yr
Y Yn
1
= (q 2 /abde, q 2 /abce, q/cd; q, p)|N | (qxr xs ; q, p)Nr
(qxr xs ; q, p)Nr +Ns r,s=1
1≤r<s≤n
n
Y 1
× 2 x /abe, q −Nr f /x , qx /c, qx /d; q, p)
r=1
(q r r r r Nr
X (q/be, q/ae, q/ab; q, p)|y|
× q |y|
0≤y≤N
(q 2 /abde, q 2 /abce, q −|N |cd; q, p)|y|
Yn Y n
|y|+yr +1
θ(q xr /abe; p) q ys θ(q yr −ys xr /xs ; p) Y (q −Ns xr /xs ; q, p)yr
×
r=1
θ(qxr /abe; p) θ(xr /xs ; p) r,s=1
(qxr /xs ; q, p)yr
1≤r<s≤n
n
Y (cxr , dxr , f xr ; q, p)yr (qxr /abe; q, p)|y|
× ,
r=1
(qxr /a, qxr /b, qxr /e; q, p)yr (q Nr +2 xr /abe; q, p)|y|

where abcdef = q |N |+2 .

In the case when n = 1, by simplification this theorem reduces to the following elliptic
analogue of Bailey’s 10 φ9 transformation formula.
Corollary 5.5 (see [22, (11.2.23)])
2 −N
12 V11 (x ; ax, bx, cx, dx, ex, f x, q ; q, p)
No.6 Z.Z. Zhang & J.L. Huang: THE Cn WP-BAILEY CHAIN 1803

(q 2 /abde, q 2 /abce, q/cd, qx2 ; q, p)N

=
(q 2 x/abe, q 2 /abcdex, qx/e, qx/d; q, p)N
× 12 V11 (qx/abe; cx, dx, f x, q/be, q/ae, q/ab, q −N ; q, p), (5.14)
where abcdef = q N +2 .

Acknowledgements We would like to thank Professor S. O. Warnaar very much for

very useful discussions and suggestions.

References
[1] Agarwal A K, Andrews G, Bressoud D. The Bailey lattice. J Indian Math Soc, 1987, 51: 57–73
[2] Andrews G E. Multiple series Rogers-Ramanujan type identities. Pacific J Math, 1984, 114: 267–283
[3] Andrews G E. Umbral calculus, Bailey chains, and pentagonal number theorems. J Comb Theory, Ser A,
2000, 91(1/2): 464–475
[4] Andrews G E. Bailey’s transform, lemma, chains and tree//Special Functions 2000: Current Perspective
and Future Directions, NATO Sci Ser. II Math Phys Chem, Tempe, AZ, vol 30. Dordrecht: Kluwer Acad
Publ, 2001: 1–22
[5] Andrews G E, Berkovich A. The WP-Bailey tree and its implications. J London Math Soc, 2002, 66(2):
529–549
[6] Andrews G E, Schilling A, Warnaar S O. An A2 Bailey lemma and Rogers-Ramanujan-type identities. J
Amer Math Soc, 1999, 12(3): 677–702
[7] Bailey W N. Identities of the Rogers-Ramanujan type. Proc London Math Soc, 1949, 50(2): 1–10
[8] Bartlett N, Warnaar S O. Hall-Littewood polynomials and characters of affine Lie algebras. Adv Math,
2015, 285: 1066–1105
[9] Berkovich A, Warnaar S O. Positivity preserving transformations for q-binomial coefficients. Trans Amer
Math Soc, 2005, 357: 2291–2351
[10] Bhatnagar G. Dn basic hypergeometric series. Ramanujan J, 1999, 3: 175–203
[11] Bhatnagar G, Milne S C. Generalized bibasic hypergeometric series, and their U (n) extensions. Adv Math,
1997, 131: 188–252
[12] Bhatnagar G, Schlosser M J. Cn and Dn very-well-poised 10 φ9 transformations. Constr Approx, 1998, 14:
531–567
[13] Bhatnagar G, Schlosser M J. Elliptic well-poised Bailey transforms and lemmas on root system.
arXiv:1704.00020
[14] Bressoud D M. Some identities for terminating q-series. Math Proc Cambridge Philos Soc, 1981, 89(2):
211–223
[15] Bressoud D M. The Bailey lattice: an introduction//Ramanujan revisited (Urbana-Champaign, III, 1987),
Boston, MA: Acdemic Press, 1988: 57–67
[16] Bressoud D M, Ismail M E H, Stanton D. Change of base in Bailey pairs. Ramanujan J, 2000, 4: 435–453
[17] Coskun H. A BCn Bailey Lemma and Generalizations of Rogers-Ramanujan Identities [D]. Texas A&M
University, 2003
[18] Coskun H. An elliptic BCn Bailey lemma, multiple Rogers-Ramanujan identities and Euler’s pentagonal
number theorems. Trans Amer Math Soc, 2008, 360(10): 5397–5433
[19] Coskun H. Multilateral basic hypergeometric summation identities and hyperoctahedral group symmetries.
Adv Appl Disc Math, 2010, 5(2): 145–157
[20] Coskun H. A multilateral Bailey lemma and multiple Andrews-Gordon identities. Ramanujan J, 2011,
26(2): 229–250
[21] Coskun H, Gustafson R A. The well-poised Macdonald functions Wλ and Jackson coefficients ωλ on
BCn //Jack, Hall-Littlewood and Macdonald Polynomials. ICMS, AMS Contemporary 417. Amer Math
Soc, 2006: 127–155
[22] Gasper G, Rahman M. Basic Hypergeometric Series. Cambridge: Cambridge University Press, 1990
[23] Jouhet F. Shifted versions of the Bailey and well-poised Bailey lemmas. Ramanujan J, 2010, 23: 315–333
[24] McLaughlin J, Sills A V, Zimmer P. Lifting Bailey pairs to WP-Bailey pairs. Discrete Math, 2009, 309:
5077–5091
1804 ACTA MATHEMATICA SCIENTIA Vol.38 Ser.B

[25] McLaughlin J, Zimmer P. Some identities between basic hypergeometric series deriving from a new Bailey-
type transformation. J Math Anal Appl, 2008, 345: 670–677
[26] McLaughlin J, Zimmer P. Some implications of the WP-Bailey tree. Adv Appl Math, 2009, 43: 162–175
[27] McLaughlin J, Zimmer P. General WP-Bailey chains. Ramanujan J, 2010, 22(1): 11–31
[28] McLaughlin J, Zimmer P. A reciprocity relation for WP-Bailey pairs. Ramanujan J, 2012, 28: 155–173
[29] Lilly G M, Milne S C. The Cl Bailey transform and Bailey lemma. Constr Approx, 1993, 9: 473–500
[30] Liu Q, Ma X. On a characteristic equation of well-poised Bailey chains. Ramanujan J, 2009, 18: 351–370
[31] Lovejoy J. A Bailey lattice. Proc Amer Math Soc, 2004, 132(5): 1507–1516
[32] Milne S C. Basic hypergeometric series very well-poised in U (n). J Math Anal Appl, 1987, 122: 223–256
[33] Milne S C. Balanced 3 φ2 summation theorems for U (n) basic hypergeometric series. Adv Math, 1997, 131:
93–187
[34] Milne S C, Lilly G M. The Al and Cl Bailey transform and lemma. Bull Amer Math Soc (NS), 1992, 26:
258–263
[35] Milne S C, Lilly G M. Consequences of the Al and Cl Bailey transform and Bailey lemma. Discrete Math,
1995, 139: 319–346
[36] Rains E. BCn -symmetric abelian functions. Duke Math J, 2006, 135(1): 99–180
[37] Rosengren H. Elliptic hypergeometric series on root systems. Adv Math, 2004, 181: 417–447
[38] Rowell M J. A new general conjugate Bailey pair. Pacific J Math, 2008, 238(2): 367–385
[39] Schilling A, Warnaar S O. A higher level Bailey lemma: proof and applications. Ramanujan J, 1998, 2:
327–349
[40] Schlosser M J. Multidimensional matrix inversions and Ar and Dr basic hypergeometric series. Ramanujan
J, 1997, 1: 243–274
[41] Schlosser M J. Multilateral inversion of Ar , Cr and Dr basic hypergeometric series. Ann Comb, 2009, 13:
341–363
[42] Singh U B. A note on a transformation of Bailey. Quart J Math Oxford Ser (2), (1994, 45(177): 111–116
[43] Slater L J. Generalized Hypergeometric Functions. Cambridge: Cambridge University Press, 1966
[44] Spiridonov V P. An elliptic incarnation of the Bailey chain. Internat Math Res Notices, 2002, 37: 1945–1977
[45] Spiridonov V P. A Bailey tree for integrals. Theor Math Phys, 2004, 139: 536–541
[46] Warnaar S O. 50 years of Bailey’s lemma//Algebraic Combinatorics and Applications (Gößweinstein, 1999).
Berlin: Springer, 2001: 333–347
[47] Warnaar S O. Summation and transformation formulas for elliptic hypergeometric series. Constr Approx,
2002, 18: 479–502
[48] Warnaar S O. Extensions of the well-poised and elliptic well-poised Bailey lemma. Indag Math (NS), 2003,
14(3/4): 571–588
[49] Warnaar S O. The Bailey lemma and Kostka polynomials. J Algebraic Combin, 2004, 20(2): 131–171
[50] Zhang Z Z. Operator identities and several U (n + 1) generalizations of the Kalnins-Miller transformations.
J Math Anal Appl, 2006, 324: 1152–1167
[51] Zhang Z Z, Liu Q Y. U (n + 1) WP-Bailey tree. Ramanujan J, 2016, 40: 447–462
[52] Zhang Z Z, Wu Y. A U (n + 1) Bailey lattice. J Math Anal Appl, 2015, 426: 747–764