Jump to ContentJump to Main Navigation
Show Summary Details
More options …

Open Mathematics

formerly Central European Journal of Mathematics

Editor-in-Chief: Vespri, Vincenzo / Marano, Salvatore Angelo

IMPACT FACTOR 2018: 0.726
5-year IMPACT FACTOR: 0.869

CiteScore 2018: 0.90

SCImago Journal Rank (SJR) 2018: 0.323
Source Normalized Impact per Paper (SNIP) 2018: 0.821

Mathematical Citation Quotient (MCQ) 2018: 0.34

ICV 2018: 152.31

Open Access
See all formats and pricing
More options …
Volume 15, Issue 1


Volume 13 (2015)

Evaluation of the convolution sum involving the sum of divisors function for 22, 44 and 52

Ebénézer Ntienjem
  • Corresponding author
  • Centre for Research in Algebra and Number Theory, School of Mathematics and Statistics, Carleton University, 1125 Colonel By Drive, Ottawa, Ontario, K1S 5B6, Canada
  • Email
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
Published Online: 2017-04-21 | DOI: https://doi.org/10.1515/math-2017-0041


The convolution sum, (l,m)N02αl+βm=nσ(l)σ(m), where αβ = 22, 44, 52, is evaluated for all natural numbers n. Modular forms are used to achieve these evaluations. Since the modular space of level 22 is contained in that of level 44, we almost completely use the basis elements of the modular space of level 44 to carry out the evaluation of the convolution sums for αβ = 22. We then use these convolution sums to determine formulae for the number of representations of a positive integer by the octonary quadratic forms a(x12+x22+x32+x42)+b(x52+x62+x72+x82), where (a, b) = (1, 11), (1, 13).

Keywords: Sums of Divisors function; Convolution Sums; Dedekind eta function; Modular Forms; Eisenstein Series; Cusp Forms; Octonary quadratic Forms; Number of Representations

MSC 2010: 11A25; 11E20; 11E25; 11F11; 11F20; 11F27

1 Introduction

Let in the sequel ℕ, ℕ0, ℤ, ℚ, ℝ and ℂ denote the sets of positive integers, non-negative integers, integers, rational numbers, real numbers and complex numbers, respectively.

Suppose that k, n∊ ℕ. Then the sum of positive divisors of n to the power of k, σk(n), is defined by σk(n)=0<d|ndk.(1) We write σ(n) as a synonym for σ1(n). For m ∉ ℕ we set σk(m) = 0.

Suppose now that α, β ∊ ℕ are such that αβ. Then the convolution sum, W(α, β) (n), is defined as follows: W(α,β)(n)=(l,m)N02αl+βm=nσ(l)σ(m).(2) We write Wβ(n) as a synonym for W(1,β) (n). Given α, β ∊ ℕ, if for all (l, m) ∊ N02 it holds that α l + β mn then we set W(α, β) (n) = 0.

For those convolution sums W(α, β)(n) that have so far been evaluated, the levels αβ are given in Table 1.

Table 1

Known convolution sums W(α,β)(n)

We discuss the evaluation of the convolution sums of level αβ = 22, 44 and αβ = 52, i.e., (α, β) = (1,22), (2, 11), (1, 44), (4, 11), (1, 52), (4, 13). Convolution sums of these levels have not been evaluated yet as one can notice from Table 1.

As an application, convolution sums are used to determine explicit formulae for the number of representations of a positive integer n by the octonary quadratic forms a(x12+x22+x32+x42)+b(x52+x62+x72+x82),(3) and c(x12+x1x2+x22+x32+x3x4+x42)+d(x52+x5x6+x62+x72+x7x8+x82),(4) respectively, where a, b, c, d ∊ ℕ.

So far known explicit formulae for the number of representations of n by the octonary form Equation 3 are referenced in Table 2.

Table 2

Known representations of n by the form Equation 3

We determine formulae for the number of representations of a positive integer n by the octonary quadratic form Equation 3 for which (a, b) = (1, 11), (1, 13). These formulae for the number of representations are also new according to Table 2.

This paper is organized in the following way. In Section 2 we discuss modular forms, briefly define eta functions and convolution sums, and prove the generalization of the extraction of the convolution sum. Our main results on the evaluation of the convolution sums are discussed in Section 3. The determination of formulae for the number of representations of a positive integer n is discussed in Section 4.

Software for symbolic scientific computation is used to obtain the results of this paper. This software comprises the open source software packages GiNaC, Maxima, REDUCE, SAGE and the commercial software package MAPLE.

2 Modular forms and convolution sums

Let ℍ be the upper half-plane, that is ℍ = {z ∊ ℂ| Im(z) > 0}, and let G = SL2(ℝ) be the group of 2 × 2-matrices abcd such that a, b, c, d ∊ ℝ and adbc = 1 hold. Let furthermore Γ = SL2(ℤ) be the full modular group which is a subgroup of SL2(ℝ). Let N ∊ ℕ. Then Γ(N)=abcdSL2(Z)|abcd1001(modN) is a subgroup of G and is called the principal congruence subgroup of level N. A subgroup H of G is called a congruence subgroup of level N if it contains Γ(N).

Relevant for our purposes is the following congruence subgroup: Γ0(N)=abcdSL2(Z)|c0(modN).

Let k, N ∊ ℕ and let Γ′ ⊆ Γ be a congruence subgroup of level N ∊ ℕ. Let k ∊ ℤ, γ∊ SL2(ℤ) and f : ℍ∪ ℚ∪{∞}→ ℂ∪ {∞}. We denote by f[γ]k the function whose value at z is (cz+d)k f(γ(z)), i.e., f[γ]k(z) = (cz+d)k f(γ(z)). The following definition is based on the textbook by N. Koblitz [22, p. 108].

Definition 2.1

Let N ∊ ℕ, k ∊ ℤ, f be a meromorphic function onand Γ′ ⊂ Γ a congruence subgroup of level N.

  1. f is called a modular function of weight k for Γ′ if

    (a1)for all γ∊ Γ′ it holds that f[γ]k = f.

    (a2)for any δ ∊ Γ it holds that f[δ]k(z) can be expressed in the form nZane2πiznN, wherein an ≠ 0 for finitely many n ∊ ℤ such that n < 0.

  2. f is called a modular form of weight k for Γ′ if

    (b1)f is a modular function of weight k for Γ′,

    (b2)f is holomorphic on ℍ,

    (b3)for all δ ∊ Γ and for all n ∊ ℤ such that n < 0 it holds that an = 0.

  3. f is called a cusp form of weight k for Γ′ if

    (c1)f is a modular form of weight k for Γ′,

    (c2)for all δ ∊ Γ it holds that a0 = 0.

For k, N ∊ ℕ, let 𝔐k0(N)) be the space of modular forms of weight k for Γ0(N), 𝔖k0(N)) be the subspace of cusp forms of weight k for Γ0(N), and 𝔈k0(N)) be the subspace of Eisenstein forms of weight k for Γ0(N). Then the decomposition of the space of modular forms as a direct sum of the space generated by the Eisenstein series and the space of cusp forms, i.e., 𝔐k0(N)) = 𝔈k0(N))⊕ 𝔖k0(N)), is well-known; see for example W. A. Stein’s book (online version) [23, p. 81].

As noted in Section 5.3 of W. A. Stein’s book [23, p. 86] if the primitive Dirichlet characters are trivial and 2 ≤ k is even, then Ek(q) = 12kBkn=1σk1(n)qn, where Bk are the Bernoulli numbers.

For the purpose of this paper we only consider trivial Dirichlet characters and 2 ≤ k even. Theorems 5.8 and 5.9 in Section 5.3 of [23, p. 86] also hold for this special case.

2.1 Eta functions

The Dedekind eta function, η(z), is defined on the upper half-plane ℍ by η(z)=e2πiz24n=1(1e2πinz). We set q = e2π iz. Then η(z)=q124n=1(1qn)=q124F(q), where F(q)=n=1(1qn).

M. Newman [24, 25] systematically used the Dedekind eta function to construct modular forms for Γ0(N). M. Newman determined when a function f(z) is a modular form for Γ0(N) by providing conditions (i)-(iv) in the following theorem. G. Ligozat [26] determined the order of vanishing of an eta function at the cusps of Γ0(N), which is condition (v) or (v′) in Theorem 2.2.

The following theorem is proved in L. J. P. Kilford’s book [27, p. 99] and G. Köhler’s book [28, p. 37]; we will apply that theorem to determine eta quotients, f(z), which belong to 𝔐k0(N)), and especially those eta quotients which are in 𝔖k0(N)).

Theorem 2.2

(M. Newman and G. Ligozat). Let N ∊ ℕ, D(N) be the set of all positive divisors of N, δ∊ D(N) and rδ∊ ℤ. Let furthermore f(z) = δD(N)ηrδ(δz) be an η-quotient. If the following five conditions are satisfied

  1. δD(N)δrδ0(mod24),

  2. δD(N)Nδrδ0(mod24) ≡(mod 24),

  3. δD(N)δrδ is a square in ℚ,

  4. 0<δD(N)rδ ≡ 0 (mod 4)

  5. for each dD(N) it holds that δD(N)gcd(δ,d)2δrδ0,

    then f(z)∊ 𝔐k0(N)), where k=12δD(N)rδ.

    Moreover, the η-quotient f(z) belongs to 𝔖k0(N)) if (v) is replaced by

    (v’) for each dD(N) it holds that δD(N)gcd(δ,d)2δrδ>0.

2.2 Convolution sums W(α, β)(n)

Recall that given α, β ∊ ℕ such that αβ, the convolution sum is defined by Equation 2.

As observed by A. Alaca et al. [11], we can assume that gcd(α, β) = 1. Let q ∊ ℂ be such that |q| < 1. Then the Eisenstein series L(q) and M(q) are defined as follows: L(q)=E2(q)=124n=1σ(n)qn,(5) M(q)=E4(q)=1+240n=1σ3(n)qn.(6) The following two relevant results are essential for the sequel of this work and are a generalization of the extraction of the convolution sum using Eisenstein forms of weight 4 for all pairs (α, β) ∊ ℕ2. Their proofs are given by E. Ntienjem [21].

Lemma 2.3

Let α, β ∊ ℕ. Then (αL(qα)βL(qβ))2M4(Γ0(αβ)).

Theorem 2.4

Let α, β ∊ ℕ be such that α and β are relatively prime and α < β. Then (αL(qα)βL(qβ))2=(αβ)2+n=1240α2σ3(nα)+240β2σ3(nβ)+48α(β6n)σ(nα)+48β(α6n)σ(nβ)1152αβW(α,β)(n)qn.(7)

3 Evaluation of the convolution sums W(α,β)(n), where αβ = 22, 44, 52

In this section, we give explicit formulae for the convolution sums W(1,22)(n), W(2,11)(n), W(1,44)(n), W(4,11)(n), W(1,52)(n) and W(4,13)(n).

3.1 Bases for 𝔈40(αβ)) and 𝔖40(αβ)) with αβ = 44, 52

We observe the following inclusion relations M4(Γ0(11))M4(Γ0(22))M4(Γ0(44))(8) M4(Γ0(13))M4(Γ0(26))M4(Γ0(52)).(9)

Therefore, it suffices to correspondingly determine the basis of the spaces 𝔐40(44)) and 𝔐40(52)), respectively.

We use the dimension formulae for the space of Eisenstein forms and the space of cusp forms in T. Miyake’s book [29, Thrm 2.5.2, p. 60] or W. A. Stein’s book [23, Prop. 6.1, p. 91] to deduce that dim(𝔈40(44))) = dim(𝔈40(52))) = 6, dim(𝔖40(44)) = 15 and dim(𝔖40(52)) = 18.

Let D(44) = {1, 2, 4, 11, 22, 44} and D(52) = {1, 2, 4, 13, 26, 52} be the sets of all positive divisors of 44 and 52, respectively.

Theorem 3.1

  1. The sets 𝔅E,44 = {M(qt)| tD(44)} and 𝔅E,52 = {M(qt)|tD(52)} are bases of 𝔈40(44)) and 𝔈40(52)), respectively

  2. Let 1 ≤ i ≤15 and 1 ≤ j ≤ 18 be positive integers.

    Let δ1D(44) and (r(i, δ1))i, δ1 be the Table 3 of the powers of η(δ1z).

    Let δ2D(52) and (r(j, δ2))j, δ2 be the Table 4 of the powers of η(δ2z).

    Table 3

    Exponents of η-functions being basis elements of 𝔖40(44))

    Table 4

    Exponents of η-functions being basis elements of 𝔖40(52))

    Let furthermore Ai(q)=δ1D(44)ηr(i,δ1)(δ1z)andBj(q)=δ2D(52)ηr(j,δ2)(δ2z) be selected elements of 𝔖40(44)) and 𝔖40(52)), respectively.

    Then the sets 𝔅S,44 = {Ai(q)| 1 ≤i≤15} and 𝔅S,52 = {Bj(q)| 1 ≤ j ≤ 18} are bases of 𝔖40(44)) and 𝔖40(52)), repectively.

  3. The sets 𝔅M,44 = 𝔅E,44 ∪ 𝔅S,44 and 𝔅M,52 = 𝔅E,52∪ 𝔅S,52 constitute bases of 𝔐40(44)) and 𝔐40(52)), respectively.

For 1 ≤ i ≤ 15 and 1 ≤ j ≤ 18 let in the sequel Ai(q) be expressed in the form n=1ai(n)qn and Bj(q) be expressed in the form n=1bj(n)qn.


We give the proof for the case αβ = 44. The case αβ = 52 is proved similarly.

  1. By Theorem 5.8 in Section 5.3 of W. A. Stein [23, p. 86] M(qt) is in 𝔐40(t)) for each t which is an element of D(44). Since 𝔈40(44)) has a finite dimension, it suffices to show that M(qt) with tD(44) are linearly independent. Suppose that xt ∈ ℂ with tD(44). We prove this by induction on the elements of the set D(44) which is assumed to be ascendantly ordered.

    The case t = 1 ∈ D(44) is obvious since comparing the coefficients of qt on both sides of the equation xtM(qt) = 0 clearly gives xt = 0.

    Suppose now that the cardinality of the set D(44) is greater than 1 and that M(qt) are linearly independent for all t|44 and tt1 for a given t1 with 1 < t1 < 44. Let C be the proper non-empty subset of D(44) which contains all positive divisors of 44 less than or equal to t1. Note that all positive divisors of t1 constitute a subset of C. Let us consider the non-empty subset C ∪ {t} of D(44), wherein t is the next ascendant element of D(44) which is greater than t1 the greatest element of the set C. Then tC{t}xtM(qt)=tCxtM(qt)+xtM(qt)=0.

    By the induction hypothesis it holds that xt = 0 for all tC. So, we obtain from the above equation that xt = 0 when we compare the coefficient of qt on both sides of the equation.

    Hence, the solution is xt = 0 for all t such that t is a positive divisor of 44. Therefore, the set 𝔅E,44 is linearly independent. Hence, the set 𝔅E,44 is a basis of 𝔈40(44)).

  2. The Ai(q) with 1 ≤ i ≤ 15 are obtained from an exhaustive search using Theorem 2.2 (i) − (v). Hence, each Ai(q) is an element of the space 𝔖40(44)).

    Since the dimension of 𝔖40(44)) is 15, it suffices to show that the set {Ai(q)| 1 ≤ i ≤ 15} is linearly independent. Suppose that xiCandi=115xiAi(q)=0. Then i=115xiAi(q)=n=1(i=115xiai(n))qn=0

    which gives the following homogeneous system of linear equations i=115ai(n)xi=0,1n15.(10)

    A simple computation using software for symbolic scientific computation shows that the determinant of the matrix of this homogeneous system of linear equations is non-zero. So, xi = 0 for all 1 ≤ i ≤ 15. Hence, the set {Ai(q)| 1 ≤ i ≤ 15} is linearly independent and therefore a basis of 𝔖40(44)).

  3. Since 𝔐40(44)) = 𝔈40(44)) ⊕ 𝔖40(44)), the result follows from (a) and (b).     □

    According to Equation 8 the basis elements Ai(q), where 1 ≤ i ≤ 5, are contained in 𝔖40(22)). The basis element A2(q) is the only element of the space 𝔖40(11)) that we are able to generate with the help of Theorem 2.2. Even though the basis element A14(q) looks like an element of 𝔖40(22)), it cannot be generated at level 22 using Theorem 2.2.

To evaluate the convolution sums W(1,22)(n) and W(2,11)(n), we determine two additional basis elements of 𝔖40(22)) which are A6(q)=η(2z)η3(11z)η5(22z)η(z)=n=1a6(n)qn,A7(q)=η9(2z)η7(11z)η5(z)η3(22z)=n=1a7(n)qn.

Due to Equation 9 the basis elements Bj(q), where 1 ≤ j ≤ 7 and j = 15, 17, belong to 𝔖40(26)). We are unable to generate any elements of the space 𝔖40(13)) using Theorem 2.2. We note that B2j(q) = Bj(q2), where 4 ≤ j ≤ 7, B16(q) = B15(q2) and B18(q) = B17(q2). Therefore, one can easily replicate the evaluation of the convolution sums W(1,26)(n) and W(2,13)(n) shown by E. Ntienjem [21].

3.2 Evaluation of W(α,β)(n) where αβ = 22, 44, 52

Lemma 3.2

We have (L(q2)22L(q22))2=441+n=1(331261σ3(n)+1267261σ3(n2)11088061σ3(n11)+655776061σ3(n22)+1209661a1(n)+4579261a2(n)+1987261a3(n)+7372861a4(n)5068861a5(n)+22176a6(n)+864a7(n))qn,(11) (2L(q2)11L(q11))2=81+n=1(1584061σ3(n)+3744061σ3(n2)+162676861σ3(n11)49420861σ3(n22)+3686461a1(n)+35740861a2(n)+116035261a3(n)+153907261a4(n)+83404861a5(n)22176a6(n)864a7(n))qn,(12) (L(q)44L(q44))2=1849+n=1(12446461σ3(n)57766233640565σ3(n2)+689863685795σ3(n4)17424061σ3(n11)+620642885795σ3(n22)+25256901125795σ3(n44)+144061a1(n)829278725795a2(n)8873455685795a3(n)16764295685795a4(n)28040071685795a5(n)+37533807365795a6(n)1335628819a7(n)+42266096645795a8(n)63360019a9(n)5273326081159a10(n)+767923219a11(n)1523174495a12(n)13107916895a13(n)+31795219a14(n)1259596895a15(n))qn,(13) (4L(q4)11L(q11))2=49+n=1(11088061σ3(n)+801218885795σ3(n2)483386885795σ3(n4)+181790461σ3(n11)984804485795σ3(n22)273208325795σ3(n44)+11088061a1(n)+1748574725795a2(n)+11694271685795a3(n)+21141895685795a4(n)+30255137285795a5(n)35110805765795a6(n)+1331827219a7(n)36417623045795a8(n)+63360019a9(n)+6639137281159a10(n)767923219a11(n)+1523174495a12(n)+13107916895a13(n)31795219a14(n)+1259596895a15(n))qn,(14) (L(q)52L(q52))2=2601+n=1(61090081243σ3(n)4565040848166064597σ3(n2)+25459241σ3(n4)73619521243σ3(n13)48295288273446064597σ3(n26)+43473830441σ3(n52)30661441243b1(n)+4981571790486064597b2(n)+9273270707046064597b3(n)4425775005606064597b4(n)85304136696486064597b5(n)101616997322886064597b6(n)103883663521243b7(n)+104083241b8(n)+7488b9(n)+329100929664147917b10(n)+27456b11(n)152492885101446064597b12(n)+17472b13(n)+4700966441b14(n)25166713896551327b15(n)+41670318268806064597b16(n)1264250239206064597b17(n)+86860841b18(n))qn,(15) (4L(q4)13L(q13))2=81+n=1(30661441243σ3(n)2400612306726064597σ3(n2)+13939241σ3(n4)+457986721243σ3(n13)539220318246064597σ3(n26)+2029017641σ3(n52)30661441243b1(n)+2127358198806064597b2(n)+2518488510246064597b3(n)4005610378086064597b4(n)51524598204006064597b5(n)54087483121926064597b6(n)54893553121243b7(n)+15033641b8(n)7488b9(n)+151016538432147917b10(n)27456b11(n)82248324316806064597b12(n)17472b13(n)54489641b14(n)11115614088551327b15(n)+20569536096006064597b16(n)647456933286064597b17(n)230441b18(n))qn.(16)


We just prove the case (4L(q4) − 11L(q11))2. The other cases are proved similarly.

It follows from Lemma 2.3 that (4L(q4) − 11L(q11))2 ∈ 𝔐40(44)). Hence, by Theorem 3.1 (c), there exist Xδ, Yj ∈ ℂ, 1 ≤ j ≤ 15 and δD(44), such that (4L(q4)11L(q11))2=δD(44)XδM(qδ)+j=115YjAj(q)=δD(44)Xδ+n=1(240δD(44)σ3(nδ)Xδ+j=1mSaj(n)Yj)qn.(17)

We equate the right hand side of Equation 17 with that of Equation 7 when setting (α, β) = (4,11) to obtain n=1(240δD(44)σ3(nδ)Xδ+j=115aj(n)Yj)qn=n=1(3840σ3(n4)+29040σ3(n11)+192(116n)σ(n4)+528(46n)σ(n11)50688W(4,11)(n))qn.

We now take the coefficients of qn for which n is in {1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,20,22,44}.

This results in a system of linear equations whose unique solution determines the values of the unknown Xδ for all δD(44) and the values of the unkown Yj for all 1 ≤ j ≤ 15. Hence, we obtain the stated result.     □

Our main result of this section is as follows.

Theorem 3.3

Let n be a positive integer Then W(1,22)(n)=171464σ3(n)1122σ3(n2)+35488σ3(n11)+125366σ3(n22)+(124188n)σ(n)+(12414n)σ(n22)212684a1(n)1595368a2(n)695368a3(n)32671a4(n)+261a5(n)78a6(n)388a7(n),(18) W(2,11)(n)=5488σ3(n)+5366σ3(n2)+1371464σ3(n11)+39122σ3(n22)+(124144n)σ(n2)+(12418n)σ(n11)16671a1(n)12415368a2(n)40295368a3(n)668671a4(n)362671a5(n)+78a6(n)+388a7(n),(19) W(1,44)(n)=13366σ3(n)+5014431784860σ3(n2)13615795σ3(n4)+55976σ3(n11)1959192720σ3(n22)+987817385σ3(n44)+(1241176n)σ(n)+(12414n)σ(n44)510736a1(n)+35993127490a2(n)+30810611019920a3(n)+6614711590a4(n)+1217017127490a5(n)3258143254980a6(n)+52738a7(n)91723363745a8(n)+2538a9(n)+208072318a10(n)30338a11(n)+601190a12(n)+258695a13(n)69209a14(n)+497190a15(n),(20) W(4,11)(n)=35976σ3(n)2529192720σ3(n2)+417817385σ3(n4)11732σ3(n11)+1554346360σ3(n22)+5395795σ3(n44)+(124144n)σ(n4)+(124116n)σ(n11)35976a1(n)75893127490a2(n)40605111019920a3(n)917617127490a4(n)1313157127490a5(n)+3047813254980a6(n)105176a7(n)+79031363745a8(n)2538a9(n)28815725498a10(n)+30338a11(n)601190a12(n)258695a13(n)+69209a14(n)497190a15(n),(21) W(1,52)(n)=971243σ3(n)+731577059582201312σ3(n2)17164σ3(n4)+589959664σ3(n13)+7739629531582201312σ3(n26)81757492σ3(n52)+(1241208n)σ(n)+(12414n)σ(n52)+31939775632b1(n)69188497095045744704b2(n)193193139737568617056b3(n)+236419605194067104b4(n)+4556844909194067104b5(n)+135706460148516776b6(n)+554934139776b7(n)139328b8(n)18b9(n)-659256671775004b10(n)1124b11(n)+203649686348516776(n)b12(n)724b13(n)3139164b14(n)+349537693458704064b15(n)55649463548516776b16+67534735194067104b17(n)2982b18(n),(22) W(4,13)(n)=31939775632σ3(n)+50012756397568617056σ3(n2)+476396σ3(n4)+24049387816σ3(n13)+11233756637568617056σ3(n26)176132132σ3(n52)+(124152n)σ(n4)+(124116n)σ(n13)+31939b775632b1(n)29546641655045744704b2(n)52468510637568617056b3(n)+83450216217568617056b4(n)+357809709752522872352b5(n)+4695094021315359044b6(n)+38120523517088b7(n)2614264b8(n)+18b9(n)78654447146150104b10(n)+1124b11(n)+428376689151892154264b12(n)+724b13(n)+4732132b14(n)+154383529458704064b15(n)5356650025946077132b16(n)+13488686117568617056b17(n)+11066b18(n).(23)


We prove the case W(4,13)(n) as the other cases are proved similarly.

We equate the right hand side of Equation 16 with that of Equation 7 when setting (α, β) = (4,13), namely n=1(3840σ3(n4)+40560σ3(n13)+192(136n)σ(n4)+624(46n)σ(n13)59904W(4,13)(n))qn=n=1(30661441243σ3(n)2400612306726064597σ3(n2)+13939241σ3(n4)+457986721243σ3(n13)539220318246064597σ3(n26)+2029017641σ3(n52)30661441243b1(n)+2127358198806064597b2(n)+2518488510246064597b3(n)4005610378086064597b4(n)51524598204006064597b5(n)54087483121926064597b6(n)54893553121243b7(n)+15033641b8(n)7488b9(n)+151016538432147917b10(n)27456b11(n)82248324316806064597b12(n)17472b13(n)54489641b14(n)11115614088551327b15(n)+20569536096006064597b16(n)647456933286064597b17(n)230441b18(n))qn.

We then solve for W(4,13)(n) to obtain the stated result.     □

4 Number of representations of a positive integer n by the qctonary quadratic form using W(α,β)(n) when αβ = 44, 52

Let n ∈ ℕ0 and the number of representations of n by the quaternary quadratic form x12+x22+x32+x42 be denoted by r4(n). That means, r4(n)=card({(x1,x2,x3,x4)Z4|m=x12+x22+x32+x42}).

We set r4(0) = 1. For all n ∈ ℕ, the following Jacobi’s identity is proved in K. S. Williams’ book [30, Thrm 9.5, p. 83] r4(n)=8σ(n)32σ(n4).(24)

Let furthermore the number of representations of n by the octonary quadratic form a(x12+x22+x32+x42)+b(x52+x62+x72+x82) be denoted by N(a,b)(n). That means, N(a,b)(n)=card({(x1,x2,x3,x4,x5,x6,x7,x8)Z8|n=a(x12+x22+x32+x42)+b(x52+x62+x72+x82)}).

We infer the following result:

Theorem 4.1

Let n ∈ ℕ and (a, b) = (1, 11), (1, 13). Then N(1,11)(n)=8σ(n)32σ(n4)+8σ(n11)32σ(n44)+64W(1,11)(n)+1024W(1,11)(n4)256(W(4,11)(n)+W(1,44)(n)),N(1,13)(n)=8σ(n)32σ(n4)+8σ(n13)32σ(n52)+64W(1,13)(n)+1024W(1,13)(n4)256(W(4,13)(n)+W(1,52)(n)).


We only prove N(1, 11)(n) since that for N(1, 13)(n) is done similarly.

It holds that N(1,11)(n)=(l,m)N02l+11m=nr4(l)r4(m)=r4(n)r4(0)+r4(0)r4(n11)+(l,m)N2l+11m=nr4(l)r4(m).

We make use of Equation 24 to derive N(1,11)(n)=8σ(n)32σ(n4)+8σ(n11)32σ(n52)+(l,m)N2l+11m=n(8σ(l)32σ(l4))(8σ(m)32σ(m4)).

We observe that (8σ(l)32σ(l4))(8σ(m)32σ(m4))=64σ(l)σ(m)256σ(l4)σ(m)256σ(l)σ(m4)+1024σ(l4)σ(m4).

The evaluation of W(1,11)(n)=(l,m)N2l+11m=nσ(l)σ(m) is shown by E. Royer [10, Thrm 1.3]. We map l to 4l to infer W(4,11)(n)=(l,m)N2l+11m=nσ(l4)σ(m)=(l,m)N24l+11m=nσ(l)σ(m).

The evaluation of W(4,11)(n) is given in Equation 21. We next map m to 4m to conclude W(1,44)(n)=(l,m)N2l+11m=nσ(l)σ(m4)=(l,m)N2l+44m=nσ(l)σ(m).

The evaluation of W(1,44)(n) is provided by Equation 20. We simultaneously map l to 4l and m to 4m to deduce (l,m)N2l+11m=nσ(l4)σ(m4)=(l,m)N2l+11m=n4σ(l)σ(m)=W(1,11)(n4).

Again, E. Royer [10, Thrm 1.3] has proved the evaluation of W(1, 11)(n).

We then put these evaluations together to obtain the stated result for N(1, 11)(n).     □


I am indebtedly thankful to the anonynuous referee for fruitful comments and suggestions on a draft of this paper.


  • [1]

    M. Besge. Extrait d’une lettre de M Besge à M Liouville. J Math Pure Appl, 1885, 7, 256. Google Scholar

  • [2]

    James Whitbread Lee Glaisher. On the square of the series in which the coefficients are the sums of the divisors of the exponents. Messenger Math, 1862, 14, 156–163.Google Scholar

  • [3]

    S. Ramanujan. On certain arithmetical functions. T Cambridge Phil Soc, 1916, 22, 159–184. Google Scholar

  • [4]

    J. G. Huard, Z. M. Ou, B. K. Spearman, and Kenneth S. Williams. Elementary evaluation of certain convolution sums involving divisor functions. Number Theory Millenium, 2002, 7, 229–274. A K Peters, Natick, MA. Google Scholar

  • [5]

    Mathieu Lemire and Kenneth S. Williams. Evaluation of two convolution sums involving the sum of divisors function. Bull Aust Math Soc, 2006, 73, 107–115. CrossrefGoogle Scholar

  • [6]

    S. Cooper and P. C. Toh. Quintic and septic Eisenstein series. Ramanujan J, 2009, 19, 163–181. Web of ScienceCrossrefGoogle Scholar

  • [7]

    Şaban Alaca and Kenneth S. Williams. Evaluation of the convolution sums l+6m=nσ(l)σ(m)and2l+3m=nσ(l)σ(m). J Number Theory, 2007, 124(2), 490–510. Google Scholar

  • [8]

    Kenneth S. Williams. The convolution sum m<n8σ(m)σ(n8m). Pac J Math, 2006, 228, 387–396. Google Scholar

  • [9]

    Kenneth S. Williams. The convolution sum m<n9σ(m)σ(n9m). Int J Number Theory, 2005, 1(2), 193–205. CrossrefGoogle Scholar

  • [10]

    Emmanuel Royer. Evaluating convolution sums of divisor function by quasimodular forms. Int J Number Theory, 2007, 3(2), 231–261. CrossrefWeb of ScienceGoogle Scholar

  • [11]

    Ayşe Alaca, Şaban Alaca, and Kenneth S. Williams. Evaluation of the convolution sums l+12m=nσ(l)σ(m)and3l+4m=nσ(l)σ(m). Adv Theor Appl Math, 2006, 1(1), 27–48. Google Scholar

  • [12]

    Ayşe Alaca, Şaban Alaca, and Kenneth S. Williams. Evaluation of the convolution sums l+18m=nσ(l)σ(m)and2l+9m=nσ(l)σ(m). Int Math Forum, 2007, 2(2), 45–68. CrossrefGoogle Scholar

  • [13]

    Ayşe Alaca, Şaban Alaca, and Kenneth S. Williams. Evaluation of the convolution sums l+24m=nσ(l)σ(m)and3l+8m=nσ(l)σ(m). Math J Okayama Univ, 2007, 49, 93–111.Google Scholar

  • [14]

    Ayşe Alaca, Şaban Alaca, and Kenneth S. Williams. The convolution sum m<n16σ(m)σ(n16m). Canad Math Bull, 2008, 51(1), 3–14. Google Scholar

  • [15]

    B. Ramakrishnan and B. Sahu. Evaluation of the convolution sums l+15m=nσ(l)σ(m)and3l+5m=nσ(l)σ(m). Int J Number Theory, 2013, 9(3), 799–809. Google Scholar

  • [16]

    Shaun Cooper and Dongxi Ye. Evaluation of the convolution sums l+20m=nσ(l)σ(m),4l+5m=nσ(l)σ(m)and2l+5m=nσ(l)σ(m). J Number Theory, 2014, 10(6), 1386–1394. Google Scholar

  • [17]

    H. H. Chan and Shaun Cooper. Powers of theta functions. Pac J Math, 2008, 235, 1–14. CrossrefWeb of ScienceGoogle Scholar

  • [18]

    E. X. W. Xia, X. L. Tian, and O. X. M. Yao. Evaluation of the convolution sum l+25m=nσ(l)σ(m). Int J Number Theory, 2014, 10(6), 1421–1430. Google Scholar

  • [19]

    Şaban Alaca and Yavuz Kesicioğlu. Evaluation of the convolution sums l+27m=nσ(l)σ(m)andl+32m=nσ(l)σ(m). Int J Number Theory, 2016, 12(1), 1–13. Google Scholar

  • [20]

    Dongxi Ye. Evaluation of the convolution sums l+36m=nσ(l)σ(m)and4l+9m=nσ(l)σ(m). Int J Number Theory, 2015, 11(1), 171–183. Google Scholar

  • [21]

    Ebénézer Ntienjem. Evaluation of the convolution sums αl+βm=nσ(l)σ(m), where (α, β) is in {(1, 14), (2, 7), (1, 26), (2, 13), (1, 28), (4, 7), (1, 30), (2, 15), (3, 10), (5, 6)}. Master’s thesis, School of Mathematics and Statistics, Carleton University, 2015. Google Scholar

  • [22]

    Neal Koblitz. Introduction to Elliptic Curves and Modular Forms, volume 97 of Graduate Texts in Mathematics. Springer Verlag, New York, 2nd edition, 1993. Google Scholar

  • [23]

    William A. Stein. Modular Forms, A Computational Approach, volume 79. American Mathematical Society, Graduate Studies in Mathematics, 2011. http://wstein.org/books/modform/modform/

  • [24]

    Morris Newman. Construction and application of a class of modular functions. Proc Lond Math Soc, 1957, 7(3), 334–350. Google Scholar

  • [25]

    Morris Newman. Construction and application of a class of modular functions II. Proc Lond Math Soc, 1959, 9(3), 373–387. Google Scholar

  • [26]

    Gérard Ligozat. Courbes modulaires de genre 1. Bull Soc Math France, 1975, 43, 5–80. Google Scholar

  • [27]

    L. J. P. Kilford. Modular forms: A classical and computational introduction. Imperial College Press, London, 2008. Google Scholar

  • [28]

    G. Köhler. Eta Products and Theta Series Identities, volume 3733 of Springer Monographs in Mathematics. Springer Verlag, Berlin Heidelberg, 2011. Google Scholar

  • [29]

    Toshitsune Miyake. Modular Forms. Springer monographs in Mathematics. Springer Verlag, New York, 1989. Google Scholar

  • [30]

    Kenneth S Williams. Number Theory in the Spirit of Liouville, volume 76 of London Mathematical Society Student Texts. Cambridge University Press, Cambridge, 2011. Google Scholar

About the article

Received: 2016-08-30

Accepted: 2017-02-28

Published Online: 2017-04-21

Citation Information: Open Mathematics, Volume 15, Issue 1, Pages 446–458, ISSN (Online) 2391-5455, DOI: https://doi.org/10.1515/math-2017-0041.

Export Citation

© 2017 Ntienjem. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License. BY-NC-ND 3.0

Citing Articles

Here you can find all Crossref-listed publications in which this article is cited. If you would like to receive automatic email messages as soon as this article is cited in other publications, simply activate the “Citation Alert” on the top of this page.

Bumkyu Cho
International Journal of Number Theory, 2019

Comments (0)

Please log in or register to comment.
Log in