- Practice
- About 22 min
- 5 practice questions
The definite integral as the limit of a sum and the fundamental theorems
Build the definite integral from Riemann sums, estimate integrals by the left-end, right-end and mid-point rules, evaluate one as a limit of a sum, and use the fundamental theorems of integral calculus.
If something looks wrong, please tell us.
By the end of this lesson you can
- Explain what a Riemann sum is and how the definite integral is its limit.
- Estimate an integral with the left-end, right-end and mid-point rules.
- Evaluate a simple definite integral directly as the limit of a sum.
- State the first and second fundamental theorems and use them to evaluate integrals.
- Explain why a definite integral need not equal a geometric area.
The textbook opens Chapter 9 with Archimedes, who found the area cut off from a parabola by a straight line by splitting it into infinitely many small pieces and adding them. That idea, slicing and summing, is exactly the definite integral. This lesson builds it carefully (Section 9.2) and then shows the short cut that makes it practical (Section 9.3).
Riemann sums
Let f be a bounded function on [a, b]. Partition the interval by points a = x₀ < x₁ < x₂ < ... < xₙ = b, and in each subinterval [xᵢ₋₁, xᵢ] choose any point ξᵢ. The sum f(ξ₁)(x₁ − x₀) + f(ξ₂)(x₂ − x₁) + ... + f(ξₙ)(xₙ − xₙ₋₁) is a Riemann sum. Each term is the signed area of a thin rectangle. If, as n → ∞ and the largest width (xᵢ − xᵢ₋₁) → 0, these sums approach a single finite value A, then A is the definite (Riemann) integral ∫ₐᵇ f(x) dx. Also ∫ₐᵃ f(x) dx = 0.
With n equal widths h = (b − a)/n, the three usual choices of ξᵢ give three rules: the left-end rule takes ξᵢ = xᵢ₋₁, the right-end rule takes ξᵢ = xᵢ, and the mid-point rule takes ξᵢ = (xᵢ₋₁ + xᵢ)/2. All three tend to the same integral as n → ∞, but for a fixed n the mid-point rule is usually the most accurate.
Worked example 1: three estimates of ∫₁² (1/x) dx
Use 4 equal subintervals, so h = 0.25 and the points are 1, 1.25, 1.5, 1.75, 2. Left-end: 0.25 [f(1) + f(1.25) + f(1.5) + f(1.75)] = 0.25 [1 + 0.8 + 0.6667 + 0.5714] ≈ 0.7595. Right-end: 0.25 [0.8 + 0.6667 + 0.5714 + 0.5] ≈ 0.6345. Mid-point: 0.25 [f(1.125) + f(1.375) + f(1.625) + f(1.875)] = 0.25 [0.8889 + 0.7273 + 0.6154 + 0.5333] ≈ 0.6912. The exact value is log 2 ≈ 0.6931. Since 1/x is decreasing, the left-end sum overestimates and the right-end sum underestimates.
The limit formula
∫ₐᵇ f(x) dx = lim (n → ∞) [(b − a)/n] Σ (r = 1 to n) f(a + r(b − a)/n). The sum may equally run from r = 0 to n − 1.
Worked example 2: an integral as a limit
Evaluate ∫₀² (x² + 1) dx as the limit of a sum. Working: a = 0, b = 2, h = 2/n, and f(2r/n) = 4r²/n² + 1. So the sum is (2/n) Σ (4r²/n² + 1) = (8/n³) Σ r² + (2/n)(n). Using Σ r² = n(n + 1)(2n + 1)/6, this is (8/6)(1 + 1/n)(2 + 1/n) + 2. As n → ∞ it tends to (8/6)(1)(2) + 2 = 8/3 + 2 = 14/3.
The fundamental theorems of integral calculus
Limits of sums are slow even for simple functions. Newton and Leibniz found the link between integration and differentiation, stated in two theorems (the textbook states them without proof):
The two fundamental theorems
-
First fundamental theorem
- f continuous on [a, b]
- F(x) = ∫ₐˣ f(u) du
- Then F′(x) = f(x)
- So F is an anti-derivative of f
-
Second fundamental theorem
- f continuous on [a, b]
- F any anti-derivative of f
- Then ∫ₐᵇ f(x) dx = F(b) − F(a)
- Written [F(x)]ₐᵇ
Because F(b) − F(a) is unchanged when a constant is added to F, no constant of integration is needed in a definite integral, and its value is unique. The variable of integration is a dummy: ∫ₐᵇ f(x) dx = ∫ₐᵇ f(u) du.
Worked example 3: using the theorems
(i) Check example 2: ∫₀² (x² + 1) dx = [x³/3 + x]₀² = 8/3 + 2 = 14/3, the same answer in one line. (ii) ∫₀^(π/2) cos x dx = [sin x]₀^(π/2) = 1 − 0 = 1. (iii) ∫₁ᵉ (1/x) dx = [log x]₁ᵉ = 1 − 0 = 1. (iv) By the first theorem, d/dx ∫₀ˣ √(1 + u²) du = √(1 + x²), with no integration needed.
Integral and area are not the same thing
If f(x) ≥ 0 on [a, b], ∫ₐᵇ f(x) dx is the area between the graph, the x-axis and the lines x = a, x = b. If f(x) ≤ 0, the integral is the negative of that area. If f changes sign, split [a, b] at the sign changes: parts above the axis count positive and parts below count negative. For example, ∫₀^(2π) sin x dx = [−cos x]₀^(2π) = −1 + 1 = 0, but the area between y = sin x and the x-axis over [0, 2π] is 2 + 2 = 4.
Key takeaways: a definite integral is the limit of Riemann sums of signed rectangle areas; the left-end, right-end and mid-point rules are three choices of sample point; ∫ₐᵇ f = lim [(b − a)/n] Σ f(a + r(b − a)/n); the second fundamental theorem gives ∫ₐᵇ f = F(b) − F(a) with no constant; a definite integral counts area below the axis as negative.
Check your understanding
-
1. The right-end rule with 2 equal subintervals for ∫₀² x² dx gives
- 2
- 5
- 8/3
- 1
Show answer
Answer: B. h = 1 and the right end points are 1 and 2: 1 × [f(1) + f(2)] = 1 + 4 = 5. (The exact value is 8/3.)
-
2. Which limit of a sum equals ∫₀¹ x dx?
- lim (1/n) Σ (r = 1 to n) r
- lim Σ (r = 1 to n) r/n
- lim (1/n) Σ (r = 1 to n) r/n
- lim n Σ (r = 1 to n) r/n
Show answer
Answer: C. Here a = 0, b = 1, so the sum is (1/n) Σ f(r/n) = (1/n) Σ r/n = n(n + 1)/(2n²) → 1/2, which is ∫₀¹ x dx.
-
3. If F(x) = ∫₂ˣ (u³ + 1) du, then F′(1) equals
- 0
- 1
- −15/4
- 2
Show answer
Answer: D. By the first fundamental theorem, F′(x) = x³ + 1, so F′(1) = 2. The lower limit 2 does not matter.
-
4. ∫₀^π sin x dx equals
- 2
- 0
- 1
- −2
Show answer
Answer: A. [−cos x]₀^π = −cos π + cos 0 = 1 + 1 = 2.
-
5. ∫₋₁² x dx equals 3/2. The area between y = x, the x-axis and the lines x = −1, x = 2 is
- 3/2
- 2
- 1/2
- 5/2
Show answer
Answer: D. Split at 0: the part below the axis has area 1/2 and the part above has area 2. Area = 1/2 + 2 = 5/2, whereas the integral is 2 − 1/2 = 3/2.
Sign in to keep track of the lessons you finish.