Chapter 4 — Integration: Riemann to Lebesgue
From Riemann's sums to Lebesgue's measure: which functions can be integrated, what happens on unbounded intervals, how much a function can wiggle, how to measure a set, and when limits and integrals may be exchanged.
Archimedes found the area of a parabolic segment by exhausting it with triangles; Newton and Leibniz discovered that areas could be computed by reversing differentiation. But neither answered the question underneath: which functions have an area under their graph, and what exactly is that area? Riemann's answer (1854) — cut the domain into small pieces, multiply the height by the width, and let the pieces shrink — works beautifully for continuous functions and even for functions with infinitely many jumps. It also has limits. It cannot handle the function that is $1$ on the rationals and $0$ on the irrationals; it needs special patching (improper integrals) for unbounded functions and infinite intervals; and a pointwise limit of Riemann integrable functions can fail to be Riemann integrable, which makes it awkward to exchange limits and integrals.
In 1902 Henri Lebesgue turned the idea round. Instead of cutting the domain into intervals, cut the range into levels and ask how large — in a new sense of "length" called measure — is the set where the function lies between two levels. This needs a theory of measure for complicated sets (the Cantor set, the rationals, sets with no length at all), and it repays the effort with powerful convergence theorems that are the working tools of probability, Fourier analysis and partial differential equations. This chapter takes you along that road: the Riemann integral and its improper extensions, functions of bounded variation (whose graphs have finite length), Lebesgue measure, and the Lebesgue integral.
- compute Darboux and Riemann sums, prove integrability with Riemann's criterion, and decide integrability in one line with Lebesgue's criterion (bounded, discontinuities of measure zero);
- evaluate limits of sums as integrals and differentiate integrals with variable limits, knowing exactly when $F'=f$;
- decide the convergence of improper integrals by power counting, comparison, and the Dirichlet and Abel tests, and evaluate integrals with the Gamma and Beta functions;
- compute total variation, recognise functions of bounded variation, and write them as differences of increasing functions;
- work with outer measure, measurable sets, null sets, the Cantor set and fat Cantor sets, and measurable functions;
- use the monotone convergence theorem, Fatou's lemma and the dominated convergence theorem, and give the standard examples where limits and integrals cannot be exchanged.
How this is examined
This chapter supplies a steady stream of CSIR NET questions. In Part B, expect limits of sums that are Riemann sums in disguise, derivatives of integrals with variable limits, a Gamma/Beta evaluation, the range of a parameter for which an improper integral converges, the total variation of a simple function, or the Lebesgue measure of a set built from intervals, rationals and Cantor-type constructions. Part C, where several options may be correct, favours statements to be sorted into true and false: which of these functions are Riemann integrable, which integrals converge absolutely or only conditionally, which functions are of bounded variation, which sets are measurable or null, and whether $\lim\int f_n=\int\lim f_n$ in a given example. The wrong options are almost always the standard counterexamples collected in the ⚠️ boxes of the five sections — the Dirichlet and Thomae functions, $\frac{\sin x}{x}$, $x\sin\frac1x$, the fat Cantor set, the moving bump $n\mathbf 1_{(0,1/n)}$. GATE MA asks similar questions in single-answer and multiple-select form.
Sections in this chapter
The Riemann Integral
Partitions and Darboux sums, Riemann's criterion, Riemann sums, integrability of continuous and monotone functions, Lebesgue's criterion, properties, mean value theorems and the fundamental theorem of calculus.
Improper Integrals
Integrals of the first and second kind, $p$-tests, comparison and limit comparison, absolute and conditional convergence, Dirichlet and Abel tests, $\int_0^\infty\frac{\sin x}x$, Gamma and Beta functions, principal values.
Functions of Bounded Variation
Total variation, monotone and Lipschitz functions, the Jordan decomposition, continuous functions that are not of bounded variation, rectifiable curves, the Riemann–Stieltjes integral.
Lebesgue Measure
Outer measure, measurable sets, σ-algebras and Borel sets, null sets, the Cantor set and fat Cantor sets, a non-measurable set, measurable functions, Egorov and Lusin.
The Lebesgue Integral and Convergence Theorems
Integrals of simple, non-negative and general measurable functions, monotone convergence, Fatou's lemma, dominated convergence, Riemann versus Lebesgue, $L^1$.
You will use the completeness of $\mathbb{R}$ and the $\varepsilon$-characterisation of $\sup$ and $\inf$ (§1.2), countable sets (§1.1), convergent and Cauchy sequences and series (§2.1, §2.3), compactness of $[a,b]$ and the intermediate and extreme value theorems (§3.1), uniform continuity and monotone functions (§3.2), and the mean value theorem (§3.3). From school calculus you need fluency with antiderivatives, substitution and integration by parts; this chapter supplies the rigour, not the techniques.