00
sin x? dx. [568
)
allies of cos x 2 and sin x 2 ,
•als of these functions.
ueut to prove that the
so, the double integrals
nfinite square (or, if we
le ratio having any value
'ecedes, the values taken
seen in a very general
let the plane of xy be
►, \/(37r), &c...., then in
e maximum (positive or
successive zones decrease,
le successive zones; the
series having no determ-
'nto squares by the lines
etween successive squares,
iths continually diminish,
f the succession of the
inution in the values of
alue of the integral for
be, and which I assume
this mode of integration
ie square into indefinitely
! interesting to carry out
lat we actually obtain a
then the quantity under
m x 2 = yt to x 2 — 277-, and
568]
NOTE ON THE INTEGRALS COS X 2 dx AND sin X 2 dx.
r •
si]
J 0
59
which, for r large, may be taken to be
-if
sin udu _ 1
V(r7r) ’ d( r7r )’
viz. r being large, we have A r differing from the above value
of the order .
V(i’7T)
by a quantity
It is obviously immaterial whether we integrate from x 2 = 0 to (?i + l)7r or to
(?i+l)7r+e, where e has any value less than 7r; for by so doing, we alter the value
of the integral by a quantity less than A n+1 , and which consequently vanishes when n
is indefinitely large. And similarly, it is immaterial whether we stop at an odd or
an even value of n.
We have therefore
v =
(»+1) ii-
sin x 2 dx = A 0 — A 1 + A 2 ... 4- (—) n A n ,
or, taking n to be odd, this is
= A 0 Ai + A 2 ... A n ,
or, say it is
= (A 0 — Aj) + (A 2 — A 3 )... + (A n _ x — A n ),
viz. n here denotes an indefinitely large odd integer.
If instead of A 0 - A x + A 2 - A 3 + &c., the signs had been all positive, then the
term A being ultimately as the series would have been divergent, and would
have had no definite sum: but with the actual alternate signs, the series is convergent,
and the sum has a determinate value. To show this more distinctly, observe that we
have
sin udy
A A -I-V— 1 if' sin(ra ' + № ) rfM --if'
A ’-' r ~ { > ’ 2 J-, -Ji-nr + u) ’ a J-.
V(r7r + U) ’
or, taking the integral from — 7r to 0 and from 0 to 7r, and in the first integral
writing — u in place of u, then
A r _j — A r = \ I sin
0
udu -I
(V (i’^T — U) \J{ r ^ + u )\
where, r being large, expanding the term in { } in ascending powers of u, then
A r _ x - A r is of the order 4: and the series (A 0 — AJ +(A 2 — A 3 ) ... + (A n _! — A n ) is
therefore convergent, and the sum as n is increased approaches a definite limit. Hence
the integral v has a definite value: and similarly, the integral u has a definite value.
8—2