RO C ZN IK I POLSKIEGO TOW ARZYSTW A MATEMATYCZNEGO Séria I: PRACE MATEMATYCZNE X V (1971)
A N N A L ES SOCIETATIS MATHEMATICAE POLONAE Series I: COMMENTATIONES MATHEMATICAE X V (1971)
G. K. D
w iv e d i(Varanasi, India)
O n a sequence oî Fourier coefficients
I .
Let
£ a nbe a given infinite series with the sequence of partial sums (sn). Let {pn} be a sequence of constants, real or complex and let ns write
P n = Po + Pi + P2+--- + Pn { P - 1 = P - 1 = 0 ) . The sequence-to-sequence transformation,
1 П
( 1 - 1 ) К = P v 8 n - v ( P n ^ Q )
n u = 0
defines the sequence {tn} of the ISTorlund means of the sequence {sn}T generated by the sequence of constants {p nj . The series У,ап or the sequence
is said to be summable by Norland means [4], or summable (N, p n) to the sum s, if lim tn = s.
71 -+CG
The conditions of regularity of the method of summability (V , p n) defined by (1.1) are
lim - -
№ -*00 P n
J > * l = 0 ( \ p n\), k=0
(1.2) lim —^ = 0,
and (1.3) as n -> oo.
If p n is real and non-negative, (1.3) is automatically satisfied and then (1.2) is the necessary and sufficient condition for the regularity of the method of summability {N , p n) [1].
In the special case in which 1
P r
and, therefore
n -\-1
logw,
as n -> oo, tn reduces to the familiar harmonic mean of {snj and if it is denoted by t'n, then the series ]?an or the sequence {$„,} is said to be summable by harmonic means or summable (H) to the sum s, if lim t'n — s [5].
n->
OOIf the method of summability (Y, p n) is superimposed on the Cesâro means of order one, another method of summability (Y, p n)G11 is ob
tained.
2. Let f(x) be a periodic function with period 2 tt and integrable in the sense of Lebesgue over an interval ( — tt , n). Let the Fourier series of f(x) be
O O O O
(2.1) |« o + (апсошх+Ьп&тпх) = Jj?An(x),
n = l n= 0
and then the conjugate series of (2.1) is
oo oo
(2.2) J^(bnc,osnx—an&innx) == ^ B n(x).
n = 1 n = l
We write
y)(t) — /(a?-f t)— f(x — t) — L (where L is a constant), t
W(t) =
J \ i p ( u ) \ d u, о
P m = P r> and Р (1Д)= Р Т, where r = [1 /t] is the integral part of Ijt.
3. In 1959, Yarshney [7] proved the following theorem:
T
h e o r e mA. I f t
J \tp(u)\du = о о
t
log7
as t -> 0, then the sequence {nBn(x)} is summable ( Y , --- 1'(71 to the
value L/ tz . I l l ^ + 1 /
In this paper the particular sequence j—— is replaced by a more general sequence {pn}, and then we give a more general condition for the (Y , р п)Сх, summability of the sequence {nJBn(x}}. In what follows {p n}
is real non-negative and non-increasing sequence such that P n oo with n.
4. We establish the following theorem:
T
h e o r e m. I f
(4.1)
S e q u e n c e o f F o u r i e r c o e f f i c i e n t s
6S
as t -> 0, where X{t) and a(t) are functions of t such that X(t),a(t) and X(t) -t/a (t) increase monotonically with t, and
(4.2) X(n)Pn = 0 [ a ( P J ]
as n-> oo, then sequence {nBn(x)} is summable {N , p n)'Cx to the value L/it..
We require following Lemmas to prove our theorem:
L
e m m al([ 2 ])(i) I f {pn} is non-negative and non-increasing, then for 0 < a < 6 < oo, 0 < i < n- and any n
ь
k = a
where A is an absolute constant.
(Ü) ( 1 1 * ) Р т < р т -
L
em m a2 . I f 0 < t < l{n , then
\Qn(t)\
1 vh / sin let cos kt
F K
2j Pn~k
k — \
k t t = 0{n),
Proof.
I QJf)
П П
\Qn(t)\ = ° \ j r y ]i> n -k № t)\ = = 0 { n ) .
L
e m m a. 3. I f 0 < t < тс, then
Proof. By Lemma 1 (i) and Abel’s transformation, we have sin kt cos kt '
1 fc=1 s
P n —k
{:
k t
sin kt t
01 P „ \ ^ J Pn~k Ы2
f c = l
P k
sin (n —k)t
(n— k)t и - tPn A П 1 =T+l j 71 — 1
V mi { n —k)t I
P k....t ~ ..., 7 -. 11 +
(n — k)t
by Lemma 1 (ii) and since [7]
n I .
^ I sin Ш
I к — 7Г + 1 • 2
P ro o f of th e Theorem . If we denote the (c, 1) transform of the sequence {nBn{x)} by tn, we have after Mohanty and Nanda [3],
t n—Ljiz = i V rBr(x)— Lj7i n j
r — 1
ТГ TZ
r f 1 smnt 1 cosnt ) 1 r
/ V{t)\ ^ dnn/2 “ 2 ' +
ТГ ["sinw# coswil
nt2 t J I еЙ+о(1),
by Eiemann-Lebesgue Theorem.
On account of the regularity of the method of summability we have to show that under our assumptions
(4.4)
7t П
r wit) v i Г sm M cos fan
— r = e(1)-
Л ft' 7, _ ? *-
0 " fr=I ns 7i -> oo. We write
Qnit) ~ ,zPn n fc=1 L
sin M cos M
M2 t }
Therefore,
TZ
I — f vWQnWdt 1 In
= f y { t) Qn( t) dt+ f y>(t)Qn(t)dt+ f ip{t)Qn{t)dt
0 1 In <5
-^1 + ^2 +-^35 say, where 0 < ô < и.
Now by lemma 2 and hypothesis (4.1),
Vя r r” 1 \пХ(п)рпЛ
h = j v(t)Qn(t)dt = 0 [n f W(t)\dt\ = о I ( jp\~ I = 0(1)
о о l- J
as n -> oo, since 7i,pn < P n.
Again by Lemma 3,
S e q u e n c e o f F o u r i e r c o e f f i c i e n t s
65
<5 <5
I 2 = f v( t ) Qn( . t ) dt = 0 ^ J
- <>Ш ' ™ т ) 1 ] + » [ / i * " > £ * ] + » [ , / i
— * 2 1 + * 2 2 + 1 2 3 , S a y .
Now
but
i *V
K f ,
= 0 ( 1 ), as n
<5
1 r P T
1™ ^~Fn / n t ) - f - n 1 In
t
k+1 /1 \
*/ \
W fк
' '
1 In.
dt
] ~ ° { k ) + 0
1 Цп)рп P ri
\Pn a(-Pn)
1ln
n —1 k+ 1
o W + i £ f у ( т ) р<«»йм’
fc=l к
Wtypjt
*(pk) ] = o{pk}, as fc -> o o ,
so
71— 1
I 22 = o ( l ) + ~ - V o { ^ } = o(l) n Ù ï
nd
n 1 I n n 1IÔ ' ' n k= 1 ' '
»-l
= o i D + o f F ^ p ^ i m ) = o(i) as w -> oo, from the hypothesis, so
(4.6) * 2 = 0 ( 1 ).
Since the method of snmmability is regular, we have
(4.7) Z3 = o(l)
as n -> oo, by Biemann-Lebesgue Theorem.
Hence collecting (4.5), (4.6) and (4.7) we get
* = o ( l ) which completes the proof.
R oczniki PTM — P r a c e M a tem a ty czn e XV 5