International Journal of Pure and Applied Mathematics ————————————————————————– Volume 52 No. 1 2009, 143-153
ON THE GROWTH OF ENTIRE ALGEBROIDAL FUNCTIONS Sanjib Kumar Datta1 § , Arindam Jha2 1 Department
of Mathematics University of North Bengal Raja Rammohunpur, District Darjeeling PIN Code 734013, West Bengal, INDIA e-mail: sanjib kr
[email protected] 2 Tarangapur N.K. High School Post: Tarangapur, District Uttar Dinajpur PIN Code 733129, West Bengal, INDIA e-mails: d
[email protected],
[email protected] Abstract: In this paper we study the comperative growth properties of entire algebroidal functions. AMS Subject Classification: 30D35, 30D30 Key Words: entire algebroidal function, composition, growth, order, lower order, lower proximate order
1. Introduction, Definitions and Notations Let F be a k-valued function defined by the following irreducible equation fk F k + fk−1 F k−1 + fk−2 F k−2 + · · · + f0 = 0, where fk 6≡ 0 and all fi (i = 0, 1, 2, · · · , k) are entire functions having no common zeros. If at least one of the fi (i = 0, 1, 2, · · · , k) is transcendental then F is called a k-valued algebroidal function. Further, if fk ≡ 1 then F is called a k-valued entire algebroidal function. In the paper we establish some results on the growth properties of composition of two k-valued entire algebroidal function. We do not explain the standard notations and definitions of the theory of entire functions as those are available in [5]. Received:
March 27, 2009
§ Correspondence address: 25, School Road,
c 2009 Academic Publications
Kalianibash, Barrackpore, P.O.: Nonachandanpukur, Dist.: 24 Parganas (North), P.S.: Titagarh, PIN Code: 743102, West Bengal, INDIA
144
S.K. Datta, A. Jha
Before starting our discussion we just mention the following well known definitions: Definition 1. The order ρf and lower order λf of an entire function f is defined as log[2] M (r, f ) log[2] M (r, f ) and λf = lim inf , ρf = lim sup r→∞ log r log r r→∞ where log[k] x = log log[k−1] x for k = 1, 2, 3, · · · and log[0] x = x.
Since T (r, f ) and log M (r, f ) are mutually replaceble in the formula for the order and lower order of f then ρf = lim sup r→∞
log T (r, f ) log T (r, f ) and λf = lim inf . r→∞ log r log r
Definition 2. A function λf (r) is called a lower proximate order of an entire function f if: (i) λf (r) is non negative and continuous for r ≥ r0 say; (ii) λf (r) is differentiable for r ≥ r0 except possibly at isolated points at which λ′f (r − 0) and λ′f (r + 0) exist; (iii) lim λf (r) = λf < ∞; r→∞
(iv) lim rλ′f (r) log r = 0; r→∞
(r,f ) = 1. (v) lim inf log λMf (r) r→∞
r
2. Some Lemmas In this section we present some lemmas which will be needed in the sequel. 2+ε ε
Lemma 1. (see [3]) Let f and g be two entire functions. If M (r, g) > |g (0)| for any ε (> 0) then T (r, f ◦ g) < (1 + ε) T (M (r, g) , f ) , In particular if g (0) = 0, then T (r, f ◦ g) < T (M (r, g) , f )
for all r > 0.
Lemma 2. (see [2]) Let g be an entire function with λg < ∞ and ai (i = 1, 2, 3, · · · , n; n ≤ ∞) are entire functions satisfying T (r, ai ) = o {T (r, g)}.
ON THE GROWTH OF ENTIRE ALGEBROIDAL FUNCTIONS
If
n P
T (r,g) r→∞ log M (r,g)
δ (ai , g) = 1 then lim
i=1
145
= π1 .
Lemma 3. Let f be an entire function. Then for δ (> 0) the function is ultimately an increasing function of r.
r λf +δ−λf (r)
Proof. Since d λf +δ−λf (r) r = λf + δ − λf (r) − rλ′f (r) log r r λf +δ−λf (r)−1 > 0, dr for all sufficiently large values of r, the lemma is proved.
3. Theorems In this section we present the main results of the paper. Theorem 1. such that
Let F and G be two k-valued entire algebroidal functions F k + fk−1 F k−1 + fk−2 F k−2 + · · · + f0 = 0
and Gk + gk−1 Gk−1 + gk−2 Gk−2 + · · · + g0 = 0 , where fi (i = 0, 1, 2, · · · , k − 1) and gi (i = 0, 1, 2, · · · , k − 1) are entire functions having no common zeros. Also let 0 < λfi ≤ ρfi < ∞, 0 < λgi ≤ ρgi < ∞, 0 < ρF < ∞ and 0 < ρG < ∞ for i = 0, 1, 2, · · · , k − 1. Then lim inf r→∞
k−1 log[2] T (r, F ◦ G) Y log[2] T (r, fi ◦ gi ) ρG lim inf ≤ r→∞ log T (r, F ) log T (r, gi ) ρF i=0
k−1 log[2] T (r, fi ◦ gi ) log[2] T (r, F ◦ G) Y lim sup . ≤ lim sup log T (r, F ) log T (r, gi ) r→∞ r→∞ i=0
Proof. We know that for r > 0 (Niino and Yang[4]) 1 r 1 M , gi + o (1) , fi . (1) T (r, fi ◦ gi ) ≥ log M 3 8 4 Since λfi and λgi are the lower orders of fi and gi respectively, for given εi and for all large values of r we get log M (r, fi ) > r λfi −εi and log M (r, gi ) > r λgi −εi ,
146
S.K. Datta, A. Jha
where 0 < εi < min {λfi , λgi }. So from (1) we get for all large values of r, λfi −εi λfi −εi 1 1 r 1 1 r ≥ M , gi + o (1) M , gi , T (r, fi ◦ gi ) ≥ 3 8 4 3 9 4 r λg −εi i , log T (r, fi ◦ gi ) ≥ O (1) + (λfi − εi ) 4 log[2] T (r, fi ◦ gi ) ≥ O (1) + (λgi − εi ) log r .
(2)
Also for a sequence of values of r tending to infinity, log T (r, gi ) < (λgi + εi ) log r.
(3)
So from (2) and (3) it follows for a sequence of values of r tending to infinity, log[2] T (r, fi ◦ gi ) O (1) + (λgi − εi ) log r > . log T (r, gi ) (λgi + εi ) log r Since εi (> 0) is arbitrary we obtain from (4) that
(4)
log[2] T (r, fi ◦ gi ) ≥ 1. log T (r, gi ) r→∞ Now taking products for i = 0, 1, 2, · · · , k − 1 it follows from (5) that lim sup
k−1 Y i=0
lim sup r→∞
(5)
log[2] T (r, fi ◦ gi ) ≥ 1. log T (r, gi )
(6)
Again in a similar manner, for a sequence of values of r tending to infinity, log[2] T (r, F ◦ G) O (1) + (ρG − ε) log r > , log T (r, F ) (ρF + ε) log r log[2] T (r, F ◦ G) ρG ≥ . log T (r, F ) ρF r→∞ Thus from (6) and (7) we obtain, lim sup
k−1 log[2] T (r, fi ◦ gi ) ρG log[2] T (r, F ◦ g) Y lim sup ≥ . lim sup log T (r, F ) log T (r, gi ) ρF r→∞ r→∞
(7)
(8)
i=0
Now by Lemma 1 we get for all sufficiently large values of r,
log T (r, fi ◦ gi ) ≤ O (1) + log T (M (r, gi ) , fi ) , log T (r, fi ◦ gi ) ≤ O (1) + (ρfi + εi ) log M (r, gi ) .
(9)
Also for a sequence of values of r tending to infinity, log M (r, gi ) < r λgi +εi .
(10)
ON THE GROWTH OF ENTIRE ALGEBROIDAL FUNCTIONS
147
So from (9) and (10) it follows for a sequence of values of r tending to infinity, log T (r, fi ◦ gi ) < O (1) + (ρfi + εi ) r λgi +εi , log[2] T (r, fi ◦ gi ) < O (1) + (λgi + εi ) log r.
(11)
Also for all large values of r and for 0 < εi < λgi , log T (r, gi ) > (λgi − εi ) log r.
(12)
Now combining (11) and (12) we get for a sequence of values of r tending to infinity, log[2] T (r, fi ◦ gi ) O (1) + (λgi + εi ) log r < , log T (r, gi ) (λgi − εi ) log r λg + εi log[2] T (r, fi ◦ gi ) ≤ i . r→∞ log T (r, gi ) λgi − εi As εi (> 0) is arbitrary it follows that lim inf
log[2] T (r, fi ◦ gi ) ≤ 1. r→∞ log T (r, gi ) Now taking products for i = 0, 1, 2, · · · , k − 1 we get from (14) that lim inf
k−1 Y i=0
lim inf r→∞
log[2] T (r, fi ◦ gi ) ≤ 1. log T (r, gi )
(13)
(14)
(15)
In a like manner for a sequence of values of r tending to infinity, log[2] T (r, F ◦ G) O (1) + (ρG + ε) log r < . log T (r, F ) (ρF − ε) log r Since εi (> 0) is arbitrary we obtain from (16) that ρG log[2] T (r, F ◦ G) ≤ . r→∞ log T (r, F ) ρF Now from (15) and (17) it follows that lim inf
lim inf r→∞
k−1 ρG log[2] T (r, fi ◦ gi ) log[2] T (r, F ◦ G) Y lim inf ≤ . r→∞ log T (r, F ) log T (r, gi ) ρF
(16)
(17)
(18)
i=0
Thus the theorem follows from (8) and (18).
Theorem 2. Let F and G be two k-valued entire algebroidal functions with F k + fk−1 F k−1 + fk−2 F k−2 + · · · + f0 = 0 and Gk + gk−1 Gk−1 + gk−2 Gk−2 + · · · + g0 = 0, where fi (i = 0, 1, 2, · · · , k − 1) and gi (i = 0, 1, 2, · · · , k − 1) are entire functions having no common zeros. Also let ρfi and λgi are both finite
148
S.K. Datta, A. Jha
for i = 0, 1, 2, · · · , k − 1. Assume that aij (i = 0, 1, 2, · · · , k − 1, j = 1, 2, 3, · · · , n; n ≤ ∞) n P are entire functions satisfying T (r, aij ) = o {T (r, gi )}. If δ (aij , gi ) = 1 for j=1
i = 0, 1, 2, · · · , k − 1 then
lim sup r→∞
k−1 Y i=0
k−1
Y log T (r, fi ◦ gi ) ρfi . ≤ πk log T (r, gi ) i=0
Proof. Taking products for i = 0, 1, 2, · · · , k − 1, from (9) we obtain for all large values of r and for arbitrary εi (> 0) that k−1 Y i=0
lim sup r→∞
k−1 Y log T (r, fi ◦ gi ) log M (r, gi ) (ρfi + εi ) ≤ + O (1) , T (r, gi ) T (r, gi )
k−1 Y i=0
i=0
log T (r, fi ◦ gi ) T (r, gi )
log M (r, gi ) T (r, gi )
(ρfi + εi ) lim sup
log M (r, gi ) . T (r, gi )
r→∞
≤
k−1 Y i=0
k−1 Y
(ρfi + εi )
≤ lim sup
i=0
r→∞
(19)
Since εi (> 0) is arbitrary, in view of Lemma 2 it follows from (19) that lim sup r→∞
This proves the theorem.
k−1 Y i=0
k−1
Y log T (r, fi ◦ gi ) ρfi . ≤ πk log T (r, gi )
(20)
i=0
Theorem 3. Let F and G be two k-valued entire algebroidal functions such that F k + fk−1 F k−1 + fk−2 F k−2 + · · · + f0 = 0 and Gk + gk−1 Gk−1 + gk−2 Gk−2 + · · · + g0 = 0, where fi and gi are entire functions having no common zeros for i = 0, 1, 2, · · · , k − 1. Also let ρF ◦G < ∞ and λG > 0. Then for any positive constant A, # " k−1 k−1 ρF ◦G k Y log[2] M (r, F ◦ G) Y log T (r, fi ◦ gi ) ≤ ρfi . π lim sup T (r, gi ) AλG r→∞ log[2] M (r A , G) i=0 i=0 Proof. From the definition of order and lower order we get for all sufficiently large values of r and for arbitrary positive ε, log[2] M (r, F ◦ G) ≤ (ρF ◦G + ε) log r
(21)
log[2] M r A , G ≥ A (λG − ε) log r.
(22)
and
ON THE GROWTH OF ENTIRE ALGEBROIDAL FUNCTIONS
149
Now from (21) and (22) it follows for all large values of r, log[2] M (r, F ◦ G) log lim sup
[2]
M (r A , G)
≤
(ρF ◦G + ε) , A (λG − ε)
log[2] M (r, F ◦ G) [2]
(r A , G)
≤
log M As ε (> 0) is arbitrary we get from above that r→∞
lim sup
log[2] M (r, F ◦ G) [2]
(r A , G)
(ρF ◦G + ε) . A (λG − ε)
≤
ρF ◦G . AλG
log M Now in view of (20) and (23) we obtain that " # k−1 log[2] M (r, F ◦ G) Y log T (r, fi ◦ gi ) lim sup T (r, gi ) r→∞ log[2] M (r A , G) r→∞
(23)
i=0
≤ lim sup
≤
log
[2]
M (r, F ◦ G)
[2]
r→∞
log
ρF ◦G k π AλG
k−1 Y
M (r A , G)
lim sup r→∞
k−1 Y i=0
log T (r, fi ◦ gi ) T (r, gi )
ρfi .
i=0
Thus the theorem is established. Theorem 4. Let F and G be two k-valued entire algebroidal functions with F k + fk−1 F k−1 + fk−2 F k−2 + · · · + f0 = 0 and Gk + gk−1 Gk−1 + gk−2 Gk−2 + · · · + g0 = 0, where fi and gi are entire functions for i = 0, 1, 2, · · · , k − 1, having no common zeros. Also let 0 < λfi ◦gi ≤ ρfi ◦gi < ∞ and 0 < λgi ≤ ρgi < ∞. Then for any positive constant A, k−1 k−1 Y log[2] M (r, fi ◦ gi ) 1 Y ρfi ◦gi lim sup ≤ Ak λg i r→∞ log[2] M (r A , gi ) i=0
i=0
and
lim inf r→∞
k−1 Y i=0
k−1 1 Y λfi ◦gi ≥ k . A ρgi log[2] M (r A , gi ) i=0
log[2] M (r, fi ◦ gi )
Proof. For arbitrary positive εi and for all sufficiently large values of r we obtain log[2] M (r, fi ◦ gi ) ≥ (λfi ◦gi − εi ) log r
(24)
log[2] M r A , gi ≤ A (ρgi + εi ) log r.
(25)
and
150
S.K. Datta, A. Jha
Taking products for i = 0, 1, 2, · · · , k − 1 from (24) and (25) we get for all large values of r k−1 Y (λf ◦g − εi ) Y log[2] M (r, fi ◦ gi ) k−1 i i ≥ . (26) [2] A A (ρ gi + εi ) i=0 i=0 log M (r , gi ) As εi (> 0) is arbitrary we obtain from (26) that lim inf r→∞
k−1 Y
log[2] M (r, fi ◦ gi ) log[2] M (r A , gi )
i=0
≥
k−1 1 Y λfi ◦gi . Ak ρgi
(27)
i=0
Again for all large values of r,
log[2] M (r, fi ◦ gi ) ≤ (ρfi ◦gi + εi ) log r
(28)
log[2] M r A , gi ≥ A (λgi − εi ) log r.
(29)
and Now taking products for i = 0, 1, 2, · · · , k − 1 , it follows from (28) and (29) for all large values of r k−1 Y (ρf ◦g + εi ) Y log[2] M (r, fi ◦ gi ) k−1 i i ≤ . (30) [2] A (λgi − εi ) log M (r A , gi ) i=0
i=0
As εi (> 0) is arbitrary it follows from (30) that lim sup r→∞
k−1 Y
k−1 1 Y ρfi ◦gi . ≤ k A λgi log[2] M (r A , gi ) i=0
log[2] M (r, fi ◦ gi )
i=0
(31)
Thus the theorem follows from (27) and (31).
Theorem 5. Let F and G be two k-valued entire algebroidal functions such that F k + fk−1 F k−1 + fk−2 F k−2 + · · · + f0 = 0 and Gk + gk−1 Gk−1 + gk−2 Gk−2 + · · · + g0 = 0, where fi and gi are entire functions having no common zeros for i = 0, 1, 2, · · · , k − 1. Also let λgi > 0 and ρfi ◦gi < ∞ then ) ( k−1 k−1 k−1 Y log[2] M (r, fi ◦ gi ) 1 Y λfi ◦gi 1 Y ρfi ◦gi , . lim inf ≤ min r→∞ log[2] M (r A , g ) Ak ρgi Ak λg i i i=0
i=0
i=0
Proof. For a sequence of values of r tending to infinity, log[2] M (r, fi ◦ gi ) ≤ (λfi ◦gi + εi ) log r
(32)
and for all large values of r, log[2] M r A , gi ≥ A (λgi − εi ) log r.
(33)
So for a sequence of values of r tending to infinity, it follows from (32) and (33)
ON THE GROWTH OF ENTIRE ALGEBROIDAL FUNCTIONS
151
that log[2] M (r, fi ◦ gi ) [2]
(r A , gi )
≤
(λfi ◦gi + εi ) . A (λgi − εi )
log M As εi (> 0) is arbitrary, we get from (34) that lim inf r→∞
log[2] M (r, fi ◦ gi ) [2]
(r A , gi )
≤
λfi ◦gi . Aλgi
log M Now taking products for i = 0, 1, 2, · · · , k − 1 we obtain from (35), k−1 1 Y λfi ◦gi . lim inf ≤ k r→∞ log [2] M (r A , g ) A λgi i i=0 i=0
k−1 Y
log[2] M (r, fi ◦ gi )
Also for a sequence of values of r tending to infinity, log[2] M r A , gi ≥ A (ρgi − εi ) log r.
(34)
(35)
(36)
(37)
From (28) and (37) it follows for a sequence of values of r tending to infinity, log[2] M (r, fi ◦ gi ) [2]
(r A , gi )
≤
(ρfi ◦gi + εi ) . A (ρgi − εi )
log M Since εi (> 0) is arbitrary we obtain from (38) that lim inf r→∞
log[2] M (r, fi ◦ gi ) [2]
(r A , gi )
≤
ρfi ◦gi . Aρgi
log M Taking products for i = 0, 1, 2, · · · , k − 1 it follows from (39) that k−1 1 Y ρfi ◦gi ≤ k . lim inf r→∞ log [2] M (r A , g ) A ρg i i i=0 i=0
k−1 Y
log[2] M (r, fi ◦ gi )
(38)
(39)
(40)
Thus the theorem follows from (36) and (40).
Theorem 6. If F and G be two k-valued entire algebroidal functions with F k + fk−1 F k−1 + fk−2 F k−2 + · · · + f0 = 0 and Gk + gk−1 Gk−1 + gk−2 Gk−2 + · · · + g0 = 0, where fi and gi are entire functions having no common zeros and ρfi ,λgi are both finite for i = 0, 1, 2, · · · , k − 1, Then ! k−1 P k−1 k−1 λgi Y Y log T (r, fi ◦ gi ) i=0 k . ρfi .2 lim inf ≤3 r→∞ T (r, gi ) i=0
i=0
Proof. If at least one of ρfi for i = 0, 1, 2, · · · , k − 1 is infinity then the result is obvious. So we suppose that ρfi < ∞ for all i = 0, 1, 2, · · · , k − 1. For any two entire functions fi and gi , the following two inequalities are well known, cf. [1], p. 18: T (r, fi ) ≤ log+ M (r, fi ) ≤ 3T (2r, fi )
(41)
152
S.K. Datta, A. Jha
and M (r, fi ◦ gi ) ≤ M (M (r, gi ) , fi ) .
(42)
For εi (> 0) we get from (41) and (42) for all large values of r that T (r, f ◦ g ) ≤ log M (M (r, g ) , f ) ≤ {M (r, g )}(ρfi +εi ) , i
i
i
i
i
log T (r, fi ◦ gi ) log M (r, gi ) ≤ (ρfi + εi ) . T (r, gi ) T (r, gi ) Since εi (> 0) is arbitrary, we obtain from above that log T (r, fi ◦ gi ) log M (r, gi ) lim inf ≤ ρfi lim inf . r→∞ r→∞ T (r, gi ) T (r, gi )
(43)
i) Since lim inf Tλ(r,g (r) = 1, for given εi (0 < εi < 1) we get for a sequence of values r→∞ r gi of r tending to infinity
T (r, gi ) < (1 + εi ) r λgi (r) .
(44)
Also for all large values of r T (r, gi ) > (1 − εi ) r λgi (r) .
(45)
Therefore, for a sequence of values of r tending to infinity we get for any δi (> 0) log M (r, gi ) T (r, gi )
≤
1 3 (1 + ε) (2r)λgi +δi . . λ g (1 − ε) (2r)λgi +δi −λgi (2r) r i (r)
≤
3 (1 + ε) λgi +δi , .2 (1 − ε)
because r λf +δ−λf (r) is ultimately an increasing function of r by Lemma 3. Since εi (> 0) and δi (> 0) are arbitrary, we get from above that λg log M (r, gi ) lim inf (46) ≤ 3.2 i . r→∞ T (r, gi ) Therefore from (43) and (46), it follows that λg log T (r, fi ◦ gi ) lim inf (47) ≤ ρfi .3.2 i . r→∞ T (r, gi ) Now taking products for i = 0, 1, 2, · · · , k − 1 we get from (47) that ! k−1 P k−1 k−1 λg Y Y i log T (r, fi ◦ gi ) i=0 k . ≤3 ρfi .2 lim inf r→∞ T (r, gi ) i=0
This proves the theorem.
i=0
ON THE GROWTH OF ENTIRE ALGEBROIDAL FUNCTIONS
153
References [1] W.K. Hayman, Meromorphic Functions, The Clarendon Press, Oxford (1964). [2] Q. Lin, C. Dai, On a conjecture of Shah concerning small functions, Kexue Tongbao, English Edition, 31, No. 4 (1986), 220-224. [3] K. Niino, N. Suita, Growth of a composite function of entire functions, Kodai Math. J., 3 (1980), 374-379. [4] K. Niino, C.C. Yang, Some growth relationships on the factors of two composite entire functions, Factorization Theory of Meromorphic Fumctions and Related Topics, Marcel Dekker, New York-Basel, (1982), 95-99. [5] G. Valiron, Lectures on the general theory of integral functions, Chelsea Publishing Company (1949).
154