JCA Journal of Complex Analysis 2314-4971 2314-4963 Hindawi Publishing Corporation 538592 10.1155/2013/538592 538592 Research Article Some New Explicit Values of Quotients of Theta-Function ϕ(q) and Applications to Ramanujan's Continued Fractions Saikia Nipen Tarkhanov Nikolai Department of Mathematics Rajiv Gandhi University Rono Hills Doimukh Arunachal Pradesh 791112 India rgu.ac.in 2013 7 3 2013 2013 14 11 2012 23 01 2013 2013 Copyright © 2013 Nipen Saikia. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

We find some new explicit values of the parameter hk,n for positive real numbers k and n involving Ramanujan's theta-function ϕ(q) and give some applications of these new values for the explicit evaluations of Ramanujan's continued fractions. In the process, we also establish two new identities for ϕ(q) by using modular equations.

1. Introduction

For q:=e2πiz, Im(z)>0, define Ramanujan's theta-function ϕ(q) as (1)ϕ(q):=n=-qn2=ϑ3(0,2z), where ϑ3 [1, page 464] is one of the classical theta-functions.

In his notebook [2, volume I, page 248], Ramanujan recorded several explicit values of theta-functions ϕ(q) and its quotients which are proved by Berndt and Chan . They also found some new explicit values. An account of these can also be found in Berndt's book . Yi  also evaluated many new values of ϕ(q) by finding explicit values of the parameters hk,n and hk,n for positive real numbers k and n which are defined by (2)hk,n:=ϕ(q)k1/4ϕ(qk),q=e-πn/k,hk,n:=ϕ(-q)k1/4ϕ(-qk),q=e-2πn/k. Yi  established several properties of these parameters and found their explicit values by appealing to transformation formulas and theta-function identities for ϕ(q). Recently, Saikia  found many new explicit values of quotients of ϕ(q) by finding explicit values of the parameter An which is a particular case of the parameter hk,n where k=4. Saikia  also established some new theta-function identities for ϕ(q).

In the sequel of the previous work, in this paper we find some new explicit values of the parameters h4,n and h2,n which are particular cases of the parameter hk,n by using some properties of hk,n established in  and two new theta-function identities for ϕ(q). In addition, we give some applications of these new values of hk,n for the explicit evaluations of Ramanujan's continued fractions c(q) and K(q) defined by (3)c(q):=11+2q1-q2+q2(1+q2)21-q6+q4(1+q4)21-q10+,|q|<1,(4)K(q):=q1/21+q+q21+q3+q41+q5+,|q|<1. The continued fraction c(q) is studied by Adiga and Anitha . For explicit evaluations of c(q), see . The continued fraction K(q) is called Ramanujan-Go¨llnitz-Gordon continued fraction [4, page 50]. For further references on K(q), see [8, 9].

In Section 2, we record some preliminary results. Section 3 is devoted to prove two new identities for theta-function ϕ(q). In Section 4, we find some new explicit values of the parameter h4,n. In Section 5, we evaluate some new values of the parameter h2,n. Finally in Section 6, we give applications of these values of h4,n and h2,n for the explicit evaluations of Ramanujan's continued fractions c(q) and K(q).

We end the introduction by defining Ramanujan's modular equation. The complete elliptic integral of the first kind K(k) is defined by (5)K(k):=0π/2dϕ1-k2sin2ϕ=π2n=0(1/2)n2(n!)2k2n=π22F1(12,12;1;k2), where 0<k<1,   2F1 denotes the ordinary or Gaussian hypergeometric function, and (6)(a)n=a(a+1)(a+2)(a+n-1). The number k is called the modulus of K, and k:=1-k2 is called the complementary modulus. Let K,K,L, and L denote the complete elliptic integrals of the first kind associated with the moduli k, k, l, and l, respectively. Suppose that the equality (7)nKK=LL holds for some positive integer n. Then a modular equation of degree n is a relation between the moduli k and l which is implied by (7).

If we set (8)q=exp(-πKK),q=exp(-πLL), we see that (7) is equivalent to the relation qn=q. Thus, a modular equation can be viewed as an identity involving theta-functions at the arguments q and qn. Ramanujan recorded his modular equations in terms of α and β, where α=k2 and β=l2. We say that β has degree n over α. The multiplier m connecting α and β is defined by m=z1/zn where zr=ϕ2(qr).

2. Preliminary Results Lemma 1 (see [<xref ref-type="bibr" rid="B10">5</xref>, page 385, Theorem 2.2]).

For positive real numbers k and n,

hk,1=1,

hk,1/n=1/hk,n.

Lemma 2 (see [<xref ref-type="bibr" rid="B10">5</xref>, page 387, Corollary 2.6]).

For positive real numbers k and n, (9)hk2,n=hk,nkhk,n/k.

Lemma 3 (see [<xref ref-type="bibr" rid="B10">5</xref>, page 392, Theorem 4.6]).

One has (10)2(h2,2nh2,4n+1h2,2nh2,4n)=h2,4nh2,n+2, for any positive real number n.

Lemma 4 (see [<xref ref-type="bibr" rid="B3">10</xref>, page 122, Entry 10 (i) and (v)]).

One has

ϕ(q)=z,

ϕ(q4)=z(1+(1-α)1/4))/2.

Lemma 5 (see [<xref ref-type="bibr" rid="B3">10</xref>, page 280-281, Entry 13(vii)]).

If β has degree 5 over α, then (11)(αβ3)1/8+{(1-α)(1-β)3}1/8=(α3β)1/8+{(1-α)3(1-β)}1/8.

Lemma 6 (see [<xref ref-type="bibr" rid="B3">10</xref>, page 314, Entry 19(i)]).

If β has degree 7 over α, then (12)(αβ)1/8+{(1-α)(1-β)}1/8=1.

3. Two New Identities for <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M97"><mml:mi>ϕ</mml:mi><mml:mo mathvariant="bold">(</mml:mo><mml:mi>q</mml:mi><mml:mo mathvariant="bold">)</mml:mo></mml:math></inline-formula>

In this section, we prove two new identities for theta-function ϕ(q) by using Ramanujan's modular equations and transformation formulas.

Theorem 7.

If P=ϕ(q)/ϕ(q4), Q=ϕ(q5)/ϕ(q20), then (13)P6-256PQ+640P2Q-640P3Q+320P4Q-70P5Q+640PQ2-1600P2Q2+1600P3Q2-785P4Q2+160P5Q2-640PQ3+1600P2Q3-1620P3Q3+800P4Q3-160P5Q3+320PQ4-785P2Q4+800P3Q4-400P4Q4+80P5Q4-70PQ5+160P2Q5-160P3Q5+80P4Q5-16P5Q5+Q6=0.

Proof.

Transcribing P and Q using Lemma 4(i) and (iv) and then simplifying, we get (14)(1-α)1/4=2P-1,(1-β)1/4=(2Q-1), where β has degree 5 over α.

Equivalently, (15)α=1-(2P-1)4,β=1-(2Q-1)4. Now by Lemma 5, we have (16)(αβ3)1/8-(α3β)1/8={(1-α)3(1-β)}1/8-{(1-α)(1-β)3}1/8. Squaring (16) and simplifying, we arrive at (17)(αβ3)1/4+(α3β)1/4=x+2(αβ)1/2, where(18)x={(1-α)3(1-β)}1/4+{(1-α)(1-β)3}1/4-2{(1-α)(1-β)}1/2. Squaring (17) and simplifying, we obtain (19)(αβ3)1/2+(α3β)1/2=y+4x(αβ)1/2, where y=x2+2αβ.

Squaring (19) and simplifying, we obtain (20)αβ3+α3β+2α2β2-y2-16x2αβ=8xy(αβ)1/2. Again squaring (20), we obtain (21)(αβ3+α3β+2α2β2-y2-16x2αβ)2-64x2y2αβ=0. Now employing (14) and (15) and factorizing using Mathematica, we deduce that (22)f(P,Q)g(P,Q)h(P,Q)=0, where (23)f(P,Q)=(P-Q)4,g(P,Q)=P6-256PQ+640P2Q-640P3Q+320P4Q-70P5Q+640PQ2-1600P2Q2+1600P3Q2-785P4Q2+160P5Q2-640PQ3+1600P2Q3-1620P3Q3+800P4Q3-160P5Q3+320PQ4-785P2Q4+800P3Q4-400P4Q4+80P5Q4-70PQ5+160P2Q5-160P3Q5+80P4Q5-16P5Q5+Q6,h(P,Q)=16P10-96P9Q-32P10Q+240P8Q2+192P9Q2+24P10Q2-320P7Q3-480P8Q3-144P9Q3-8P10Q3+256P6Q4+608P7Q4+384P8Q4+40P9Q4+P10Q4-448P5Q5+352P6Q5-1264P7Q5+248P8Q5-70P9Q5+256P4Q6+352P5Q6-880P6Q6+1640P7Q6-785P8Q6+160P9Q6-320P3Q7+608P4Q7-1264P5Q7+1640P6Q7-1620P7Q7+800P8Q7-160P9Q7+240P2Q8-480P3Q8+384P4Q8+248P5Q8-785P6Q8+800P7Q8-400P8Q8+80P9Q8-96PQ9+192P2Q9-144P3Q9+40P4Q9-70P5Q9+160P6Q9-160P7Q9+80P8Q9-16P9Q9+16Q10-32PQ10+24P2Q10-8P3Q10+P4Q10. By examining the behavior of the first factor f(P,Q) and the last factor h(P,Q) of the left-hand side of (22) near q=0, it can be seen that there is a neighborhood about the origin, where these factors are not zero. Then the second factor g(P,Q) is zero in this neighborhood. By the identity theorem, this factor is identically zero. Hence, we complete the proof.

Theorem 8.

If P=ϕ(q)/ϕ(q4), Q=ϕ(q7)/ϕ(q28), then (24)P8-4096PQ+14336P2Q-21504P3Q+17920P4Q-8736P5Q+2352P6Q-280P7Q+14336PQ2-51968P2Q2+80640P3Q2-69440P4Q2+35056P5Q2-9772P6Q2+1176P7Q2-21504PQ3+80640P2Q3-129472P3Q3+115360P4Q3-60424P5Q3+17528P6Q3-2184P7Q3+17920PQ4-69440P2Q4+115360P3Q4-106330P4Q4+57680P5Q4-17360P6Q4+2240P7Q4-8736PQ5+35056P2Q5-60424P3Q5+57680P4Q5-32368P5Q5+10080P6Q5-1344P7Q5+2352PQ6-9772P2Q6+17528P3Q6-17360P4Q6+10080P5Q6-3248P6Q6+448P7Q6-280PQ7+1176P2Q7-2184P3Q7+2240P4Q7-1344P5Q7+448P6Q7-64P7Q7+Q8=0.

Proof.

Transcribing P and Q using Lemma 4(i) and (iv) and then simplifying, we obtain (25)(1-α)1/4=2P-1,(1-β)1/4=(2Q-1), where β has degree 7 over α.

Equivalently, (26)α=1-(2P-1)4,β=1-(2Q-1)4. Now by Lemma 6, we have (27){(1-α)(1-β)}1/8=1-(αβ)1/8. Squaring (27) and simplifying, we obtain (28)x-(αβ)1/4=-2(αβ)1/8, where x={(1-α)(1-β)}1/4-1.

Squaring (28) and simplifying, we obtain (29)x2+(αβ)1/2=(4+2x)(αβ)1/4. Squaring (29) and simplifying, we obtain (30)x4+αβ=((4+2x)2-2x2)(αβ)1/2. Again, squaring (30), we arrive at (31)(x4+αβ)2=((4+2x)2-2x2)2(αβ). Employing (25) and (26) in (31) and simplifying with the help of Mathematica, we complete the proof.

4. New Values of <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M136"><mml:mrow><mml:msub><mml:mrow><mml:mi>h</mml:mi></mml:mrow><mml:mrow><mml:mn>4</mml:mn><mml:mo mathvariant="bold">,</mml:mo><mml:mi>n</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>

In this section, we find some new values of h4,n and h2,n by using theta-function identities proved in Section 3 and the properties of hk,n listed in Lemmas 1 and 2. We begin with following remarks.

Remark 9.

The values of hk,n are real and hk,n<1 for all n>1 if k>1. We also note that the values of hk,n decrease as n increases when k>1. In view of this, in the following theorem we have h4,4>h4,5>h4,7>h4,8>h4,25>h4,49 where h4,4 and h4,8 are evaluated in [6, Theorem 4.3]. Yi  also evaluated the value of h4,4.

Theorem 10.

One has

h4,5=(22+105-2-10-(2+10-22+105)2-4)/2,

h4,1/5=(22+105-2-10+(2+10-22+105)2-4)/2,

h4,25=114-802+25965-183602-(114-802+25965-183602)2-1,

h4,1/25=114-802+25965-183602+(114-802+25965-183602)2-1,

h4,7=(12-72-7(34-242))/2,

h4,1/7=(12-72+7(34-242))/2,

h4,49=a+b--1+(a+b)2,

h4,1/49=a+b+-1+(a+b)2,

where a=1317-9312+3470684-24541442 and b=6940535 − 49077002 + 13061331967/(123953-876482) − 92357564014/(123953-876482).

Proof.

For (i) and (ii), setting k=4 and employing the definition of hk,n in Theorem 7, we get (32)P=2h4,n,Q=2h4,25n. Setting n=1/5 in (32), applying to (13), and then simplifying using Lemma 1(ii), we obtain (33)(h4,56+h4,5-6)-70(h4,54+h4,5-4)+3202(h4,53+h4,5-3)-1425(h4,52+h4,5-2)+19202(h4,5+h4,5-1)-3348=0. Equivalently, (34)A6-76A4+3202A3-1136A2+9602A-640=0, where (35)A:=h4,5+h4,5-1. Solving (34) for A and noting that A has positive real value greater than 1, we obtain (36)A=-2-10+22+105. Invoking (36) in (35), solving for h4,5, and using the fact in Remark 9, we complete the proof of (i). Noting h4,1/5=1/h4,5 from Lemma 1(ii), we arrive at (ii).

For proofs of (iii) and (iv), we set n=1 in (32), applying to (13) and then simplifying using Lemma 1(i), we arrive at (37)(h4,252+h4,25-2)-(456-3202)(h4,25+h4,25-1)-(674-8402)=0.

Equivalently, (38)B2+8(-57+402)B+(-674+8402)=0, where (39)B:=h4,25+h4,25-1. Solving (38) for B and noting that B has positive real value greater than 1, we obtain (40)B=2(114-802+5(5193-36722)). Invoking (40) in (39), solving for h4,25, and using the fact in Remark 9, we complete the proof of (iii). Noting h4,1/25=1/h4,25 from Lemma 1(iv), we prove (ii).

To prove (v) and (vi), applying the definition of h4,n in Theorem 8, we get (41)P=2h4,n,Q=2h4,49n. Setting n=1/7 in (41), applying in (24), and then simplifying using Lemma 1(ii), we arrive at (42)h4,716-280h4,714+23522h4,713-18508h4,712+440162h4,711-140616h4,710+1592642h4,79-262810h4,78+1592642h4,77-140616h4,76+440162h4,75-18508h4,74+23522h4,73-280h4,72+1=0. Dividing (42) by h4,78 and simplifying, we get (43)(h4,78+h4,7-8)-280(h4,76+h4,7-6)+23522(h4,75+h4,7-5)-18508(h4,74+h4,7-4)+440162(h4,73+h4,7-3)-140616(h4,72+h4,7-2)+1592642(h4,7+h4,7-1)-262810=0. Equivalently, (44)L8-288L6+23522L5-16808L4+322562L3-69120L2+389762L-18032=0, where (45)L:=h4,7+h4,7-1. Solving (44) by using Mathematica and noting that L has positive real value greater that 1 satisfying the fact in Remark 9, we obtain (46)L=12-72. Employing (46) in (45), solving for h4,7, and using the fact in Remark 9, we complete the proof of (v). Noting h4,1/7=1/h4,7, we arrive at (vi).

For proofs of (vii) and (viii), setting n=1 in (41), applying in (24), and simplifying using Lemma 1(ii), we arrive at (47)h4,494+h4,49-4+4(-1317+9312)(h4,493+h4,49-3)+28(-2053+14522)(h4,492+h4,49-2)+56(-2409+17032)(h4,49+h4,49-1)+210(-837+5922)=0. Equivalently,(48)D4+8(-1317+9312)D3+16(-3593+25412)D2+64(-1614+11412)D+128(-475+3362)=0, where (49)D=h4,49+h4,49-1. Solving (48) for D and noting that D has positive real value greater than 1, we obtain (50)D=2(a+b), where a=1317-9312+3470684-24541442 and b=6940535 − 49077002 + 13061331967/(123953-876482) − 92357564014/(123953-876482).

Invoking (50) in (49), solving for h4,49, and using the fact in Remark 9, we arrive at (vi). Noting h4,1/49=1/h4,49 from Lemma 1(ii), we complete the proof of (vii).

5. New Values of <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M215"><mml:mrow><mml:msub><mml:mrow><mml:mi>h</mml:mi></mml:mrow><mml:mrow><mml:mn>2</mml:mn><mml:mo mathvariant="bold">,</mml:mo><mml:mi>n</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>

In this section, we find some new values of the parameter h2,n by using the values of h4,n evaluated in Section 4 and in .

Theorem 11.

One has

h2,6=82-10+(2-2)18-122,

h2,3/2=(2-2+18-122)/(282-10+(2-2)18-122),

h1/6=1/82-10+(2-2)18-122,

h2,2/3=282-10+(2-2)18-122/(2-2+18-122).

Proof.

Setting k=2 and n=3 in Lemma 2, we deduce that (51)h4,3=h2,6h2,3/2. From [6, page 174, Theorem 4.3(vii)], we have (52)h4,3=(2-2+18-122)2. Combining (51) and (52), we obtain (53)h2,6h2,3/2=(2-2+18-122)2. Next, setting n=1/6 in Lemma 3 and simplifying using Lemma 1(ii), we obtain (54)2(h2,6h2,3/2+(h2,6h2,3/2)-1)=(h2,6h2,3/2)+2. Invoking (53) in (54) and simplifying, we deduce that (55)(h2,6h2,3/2)=2(82-10+(2-2)18-122)2-2+18-122. Multiplying (53) and (55) and simplifying, we complete the proof of (i). Dividing (53) by (55) and simplifying, we arrive at (ii). (iii) and (iv) follow from (i) and (ii), respectively, and Lemma 1(ii).

The proofs of Theorems 1215 are identical to the proof of Theorem 11. So we omit details and give only references of the required results to prove them.

Theorem 12.

One has

h2,10=(4+c2)2-4c/2,

h2,5/2=c/(4+c2)2-4c,

h2,1/10=2/(4+c2)2-4c,

h2,2/5=(4+c2)2-4c/c,

where c=22+105-2-10-(2+10-22+105)2-4.

Proof of Theorem 12 follows from Theorem 10(i), Lemma 2 with k=2 and n=5, Lemma 1(ii), and Lemma 3 with n=1/10.

Theorem 13.

One has

h2,50=-2d+2(d2+1),

h2,25/2=d/-2d+2(d2+1),

h2,1/50=1/-2d+2(d2+1),

h2,2/25=-2d+2(d2+1)/d,

where d=114-802+25965-183602-(114-802+25965-183602)2-1.

To prove Theorem 13, we use Theorem 10(iii), Lemma 2 with k=2 and n=25, Lemma 1(ii), and Lemma 3 with n=1/50.

Theorem 14.

One has

h2,14=(-8+62)(12-72-238-1682),

h2,7/2=12-72-238-1682/2-8+62,

h2,1/14=1/(-8+62)(12-72-238-1682),

h2,2/7=2-8+62/12-72-238-1682.

We employ Theorem 10(v), Lemma 2 with k=2 and n=7, Lemma 1(ii), and Lemma 3 with n=1/14 to prove Theorem 14.

Theorem 15.

One has

h2,98=2(a+b-(a+b)2-1)(2a+2b-1),

h2,49/2=(a+b-(a+b)2-1)/2(2a+2b-1),

h2,1/98=1/2(a+b-(a+b)2-1)(2a+2b-1),

h2,2/49=2(2a+2b-1)/(a+b-(a+b)2-1),

where a and b are given in Theorem 10(viii).

Proof follows from Theorem 10(vii), Lemma 2 with k=2 and n=49, Lemma 1(ii), and Lemma 3 with n=1/98.

6. Applications of <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M260"><mml:mrow><mml:msub><mml:mrow><mml:mi>h</mml:mi></mml:mrow><mml:mrow><mml:mn>4</mml:mn><mml:mo mathvariant="bold">,</mml:mo><mml:mi>n</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M261"><mml:mrow><mml:msub><mml:mrow><mml:mi>h</mml:mi></mml:mrow><mml:mrow><mml:mn>2</mml:mn><mml:mo mathvariant="bold">,</mml:mo><mml:mi>n</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>

In this section, we use the new values of the parameters h4,n and h2,n to find explicit values of continued fractions c(q) and K(q) defined in (3) and (4), respectively.

The parameter h4,n is useful in finding explicit values of the continued fraction c(q). If we know values of the parameter h4,n for any positive real number n, then explicit values of c(e-πn/2) can be calculated by appealing to the following theorem.

Theorem 16 (see [<xref ref-type="bibr" rid="B8">6</xref>, page 177, Theorem 5.1]).

One has (56)c(e-πn/2)=12h4,n, where n is any positive real number.

For example, employing the value of h4,5 from Theorem 10(i) in Theorem 16, we obtain (57)c(e-π5/2)=2×((2+10-22+105)2-422+105-2-10-(2+10-22+105)2-4)-1. Similarly, we can find new values of c(e-π/25), c(e-5π/2), c(e-π/10), c(e-π7/2), c(e-π/27),  c(e-7π/2), and c(e-π/14) by employing the values of h4,n from Theorem 10(ii)–(viii), respectively, in Theorem 16. Since it is a routine calculation, we omit details.

Next, the parameter h2,n is connected to continued fraction K(q) by the following theorem.

Theorem 17 (see [<xref ref-type="bibr" rid="B2">8</xref>, page 281, Theorem 4.1]).

For any positive real number n, one has (58)K2(e-πn/2)=21/4h2,n-121/4h2,n+1.

From Theorem 17, we note that if the values of h2,n are known, then the values of K(e-πn/2) can easily be evaluated. For example, using the value of h2,6 from Theorem 11(i) in Theorem 17, we evaluate (59)K(e-π6/2)=21/482-10+(2-2)18-122-121/482-10+(2-2)18-122+1.

Similarly, we can evaluate new values of K(e-πn/2) for n = 3/2, 1/6, 2/3, 10, 5/2, 1/10, 2/5, 50, 25/2, 1/50, 2/25, 14, 7/2, 1/14, 2/7, 98, 49/2, and 2/49 by using Theorem 17 and the values of h2,n evaluated in Theorems 1115.

Acknowledgment

The author is thankful to the University Grants Commission, New Delhi, India for partially supporting the research work under the Grant no. F. No. 41-1394/2012(SR).

Whittaker E. T. Watson G. N. A Course of Modern Analysis 1996 Cambridge, UK Cambridge University Press Cambridge Mathematical Library Indian edition is published by Universal Book Stall, New Delhi, India, 1991 MR1424469 Ramanujan S. Notebooks (2 Volumes) 1957 Mumbai, India Tata Institute of Fundamental Research MR0099904 Berndt B. C. Chan H. H. Ramanujan's explicit values for the classical theta-function Mathematika 1995 42 2 278 294 10.1112/S0025579300014595 MR1376728 ZBL0862.33016 Berndt B. C. Ramanujan's Notebooks, Part V 1998 New York, NY, USA Springer 10.1007/978-1-4612-1624-7 MR1486573 Yi J. Theta-function identities and the explicit formulas for theta-function and their applications Journal of Mathematical Analysis and Applications 2004 292 2 381 400 10.1016/j.jmaa.2003.12.009 MR2047619 ZBL1060.11027 Saikia N. Some new identities for Ramanujan's theta-function ϕ(q) and applications Far East Journal of Mathematical Sciences 2010 44 2 167 179 MR2757261 ZBL1210.33022 Adiga C. Anitha N. A note on a continued fraction of Ramanujan Bulletin of the Australian Mathematical Society 2004 70 3 489 497 10.1017/S0004972700034742 MR2103981 ZBL1062.33016 Baruah N. D. Saikia N. Explicit evaluations of Ramanujan-Göllnitz-Gordon continued fraction Monatshefte für Mathematik 2008 154 4 271 288 10.1007/s00605-007-0515-z MR2425757 ZBL1170.11002 Chan H. H. Huang S.-S. On the Ramanujan-Göllnitz-Gordon continued fraction The Ramanujan Journal 1997 1 1 75 90 10.1023/A:1009767205471 MR1607529 ZBL0905.11008 Berndt B. C. Ramanujan's Notebooks, Part III 1991 New York, NY, USA Springer 10.1007/978-1-4612-0965-2 MR1117903