MATH 554 - INTEGRATION
Handout #9 - 11/20/97


Defn.  A collection of n+1 distinct points of the interval [a,b]

displaymath289

is called a partition of the interval.   In this case, we define the norm of the partition by

displaymath291

where tex2html_wrap_inline293 is the length of the i-th subinterval tex2html_wrap_inline297 .

Defn.  For a given partition P, we define the Riemann upper sum of a function f  by

displaymath303

where tex2html_wrap_inline305 denotes the supremum of f over each of the subintervals tex2html_wrap_inline297 . Similarly, we define the Riemann lower sum of a function f  by

displaymath313

where tex2html_wrap_inline315 denotes the infimum of f over each of the subintervals tex2html_wrap_inline297 . Since tex2html_wrap_inline321 , we note that

displaymath323

for any partition P.

Defn.  Suppose tex2html_wrap_inline327 are both partitions of [a,b], then tex2html_wrap_inline331 is called a refinement of tex2html_wrap_inline333 , denoted by

displaymath335

if as sets tex2html_wrap_inline337 .

Note.  If  tex2html_wrap_inline339 , it follows that tex2html_wrap_inline341  since each of the subintervals formed by tex2html_wrap_inline331 is contained in a subinterval which arises from tex2html_wrap_inline333 .

Lemma.  If tex2html_wrap_inline339 , then

displaymath349

and

displaymath351

Proof.  Suppose first that tex2html_wrap_inline333 is a partition of [a,b] and that tex2html_wrap_inline331 is the partition obtained from tex2html_wrap_inline333 by adding an additional point z. The general case follows by induction, adding one point at at time. In particular, we let

displaymath363

and

displaymath365

for some fixed i. We focus on the upper Riemann sum for these two partitions, noting that the inequality for the lower sums follows similarly. Observe that

displaymath369

and

displaymath371

where tex2html_wrap_inline373 and tex2html_wrap_inline375 . It then follows that tex2html_wrap_inline377 since

displaymath379

Defn.  If tex2html_wrap_inline333 and tex2html_wrap_inline331 are arbitrary partitions of [a,b], then the common refinement of tex2html_wrap_inline333 and tex2html_wrap_inline331 is the formal union of the two.

Corollary.  Suppose tex2html_wrap_inline333 and tex2html_wrap_inline331 are arbitrary partitions of [a,b], then

displaymath397

Proof.  Let P be the common refinement of tex2html_wrap_inline333 and tex2html_wrap_inline331 , then

displaymath405

Defn.  The lower Riemann integral of f over [a,b] is defined to be

displaymath411

Similarly, the upper Riemann integral of f over [a,b] is defined to be

displaymath417

By the definitions of least upper bound and greatest lower bound, it is evident that for any function f  there holds

displaymath421

Defn.  A function f  is Riemann integrable over [a,b] if the upper and lower Riemann integrals coincide. We denote this common value by tex2html_wrap_inline427 .

Examples: 

Theorem.  A necessary and sufficient condition for f  to be Riemann integrable is    given tex2html_wrap_inline437 , there exists a partition P of [a,b] such that

displaymath443

Proof.  First we show that (*) is a sufficient condition. This follows immediately, since for each tex2html_wrap_inline437 that there is a partition P such that (*) holds,

displaymath449

Since tex2html_wrap_inline437 was arbitrary, then the upper and lower Riemann integrals of f must coincide.

To prove that (*) is a necessary condition for f to be Riemann integrable, we let tex2html_wrap_inline457 By the definition of the upper Riemann integral as a infimum of upper sums, we can find a partition tex2html_wrap_inline333 of [a,b] such that

displaymath463

Similarly, we have

displaymath465

Let P be a common refinement of tex2html_wrap_inline333 and tex2html_wrap_inline331 , then subtracting the two previous inequalities implies,

displaymath473

Defn.  A Riemann sum for f  for a partition P of an interval [a,b] is defined by

displaymath481

where the tex2html_wrap_inline483 , satisfying tex2html_wrap_inline485 ), are arbitrary.

Corollary.  Suppose that f  is Riemann integrable on [a,b], then there is a unique number tex2html_wrap_inline491 (tex2html_wrap_inline493 such that for every tex2html_wrap_inline437 there exists a partition P of [a,b] such that if tex2html_wrap_inline501 , then

displaymath503

where tex2html_wrap_inline505 is any Riemann sum of f  for the partition tex2html_wrap_inline333 . In this sense, we can interpret

displaymath511

although we would actually need to show a little more to be entirely correct.

Proof.  Since tex2html_wrap_inline513 for all partitions, we see that parts i.) and ii.) follow from the definition of the Riemann integral. To see part iii.), we observe that tex2html_wrap_inline515 and hence that

displaymath517

But we also know that both

displaymath519

and condition (*) hold, from which part iii.) follows.    tex2html_wrap_inline521

Theorem.  If f  is continuous on [a,b], then f  is Riemann-integrable on [a,b].

Proof.  We use the condition (*) to prove that f is Riemann-integrable. If tex2html_wrap_inline437 , we set tex2html_wrap_inline535 . Since f is continuous on [a,b], f is uniformly continuous. Hence there is a tex2html_wrap_inline543 such that tex2html_wrap_inline545 if tex2html_wrap_inline547 . Suppose that tex2html_wrap_inline549 , then it follows that tex2html_wrap_inline551 . Hence

displaymath553

Theorem.  If f  is monotone on [a,b], then f  is Riemann-integrable on [a,b].

Proof.  If f is constant, then we are done. We prove the case for f monotone increasing. The case for monotone decreasing is similiar. We again use the condition (*) to prove that f is Riemann-integrable. If tex2html_wrap_inline437 , we set tex2html_wrap_inline571 and consider any partition P with tex2html_wrap_inline549 . Since f is monotone increasing on [a,b], then tex2html_wrap_inline581 and tex2html_wrap_inline583 . Hence

displaymath585

Theorem.  (Properties of the Riemann Integral) Suppose that f and g are Riemann integrable and k is a real number, then

i.) tex2html_wrap_inline593
ii.) tex2html_wrap_inline595
iii.) tex2html_wrap_inline597 implies tex2html_wrap_inline599 .
iv.) tex2html_wrap_inline601

Proof.  To prove part i.), we observe that in case tex2html_wrap_inline603 , then tex2html_wrap_inline605 and tex2html_wrap_inline607 . Hence U(P,kf)=kU(P,f) and L(P,kf)=kL(P,f). In the case that k;SPMlt;0, then tex2html_wrap_inline615 and tex2html_wrap_inline617 . It follows in this case that U(P,kf)=kL(P,f) and L(P,kf)=kU(P,f) and so

displaymath623

displaymath625

To prove property ii.) we notice that tex2html_wrap_inline627 and tex2html_wrap_inline629 for any interval I (for example, tex2html_wrap_inline633 ). Hence,

displaymath635

Let tex2html_wrap_inline437 , then since f and g are Riemann integrable, there exist partitions tex2html_wrap_inline327 such that

displaymath645

If we let P be a common refinement of tex2html_wrap_inline333 and tex2html_wrap_inline331 , then by combining inequalities (1) and (2), we see that see that

displaymath653

Property iii.) follows directly from the definition of the upper and lower integrals using, for example, the inequality tex2html_wrap_inline655 .

Property iv.) is proved by applying property iii.) to the inequality

displaymath657

from which it follows that tex2html_wrap_inline659 . But this inequality implies property iv.).    tex2html_wrap_inline521

Defn.  We extend the definition of the integral to include general limits of integration which are consistent with our earlier definition.

  1. tex2html_wrap_inline663 .
  2. tex2html_wrap_inline665 .

Theorem.  If f is Riemann integrable on [a,b], then it is Riemann integrable on each subinterval tex2html_wrap_inline671 . Moreover, if tex2html_wrap_inline673 , then

displaymath675

Proof.  We show first that condition (*) holds for the interval [c,d]. Suppose tex2html_wrap_inline437, then by (*) applied to f over the interval [a,b], we have that there exists a partition P of [a,b] such that condition (*) holds. Let tex2html_wrap_inline689 be the refinement obtained from P which contains the points c and d. Let tex2html_wrap_inline697 be the partition obtained by restricting the partition tex2html_wrap_inline689 to the interval [c,d], then

displaymath703

and so f is Riemann integrable over [c,d].

To prove the identity (3), we use the fact that condition (*) holds when f is Riemann integrable. Let tex2html_wrap_inline437 , then for tex2html_wrap_inline713 , we may apply (*) to each of the intervals I=[a,b], [a,c] and [c,b], respectively, to obtain partitions tex2html_wrap_inline719 which satisfy

displaymath721

We let P be the partition of [a,b] formed by the union of the two partitions tex2html_wrap_inline727 , and tex2html_wrap_inline689 be the common refinement of P and tex2html_wrap_inline733 . Observing that

displaymath735

we can combine with inequality (4) to obtain

displaymath737

Since tex2html_wrap_inline437 was arbitrary, then equality (3) must hold.    tex2html_wrap_inline521

Theorem.  (Intermediate Value Theorem for Integrals) If f is continuous on [a,b], then there exists tex2html_wrap_inline747 between a and b such that

displaymath753

Proof.  Since f is continuous on [a,b] and for tex2html_wrap_inline759 there holds

displaymath761

then by the Intermediate Value Theorem for continuous functions, there exists a tex2html_wrap_inline763 such that tex2html_wrap_inline765 .    tex2html_wrap_inline521

Theorem.  (Fundamental Theorem of Calculus, Part I. Derivative of an Integral) Suppose that f  is continuous on [a,b] and set tex2html_wrap_inline773 , then F is differentiable and F'(x) = f(x) for a<x<b.

Proof.  Notice that

displaymath781

for some tex2html_wrap_inline747 between tex2html_wrap_inline785 and tex2html_wrap_inline787 . Hence, as tex2html_wrap_inline789 , then tex2html_wrap_inline791 converges to tex2html_wrap_inline785 and so the displayed difference quotient has a limit of tex2html_wrap_inline795 as tex2html_wrap_inline789 .    tex2html_wrap_inline521

Theorem.  (Fundamental Theorem of Calculus, Part II. Integral of a Derivative) Suppose that F is function with a continuous derivative on [a,b], then

displaymath805 .

Proof.  Define tex2html_wrap_inline807 , and set H:=F-G. Since the derivative of H is identically zero by Part I of the Fundamental Theorem of Calculus, then the Mean Value Theorem implies that H(b) = H(a). Expressing this in terms of F and G gives

displaymath819

which establishes the theorem.    tex2html_wrap_inline521


Robert Sharpley
Fri Jan 9 23:07:34 EST 1998