Master the fundamental properties that make limits computable.
The first fundamental property: if a limit exists, it is unique. A function cannot approach two different values at the same point.
If and , then .
Suppose . Let .
By definitions, such that:
Let . For :
Contradiction! So . ∎
Uniqueness lets us write "THE limit" and use limits in equations unambiguously.
If a limit exists, the function cannot "blow up" near the limit point.
If , then f is locally bounded:
Take . : .
By triangle inequality: .
So works. ∎
has limit 1 at x = 1, but is unbounded on (0, ∞).
If and g is bounded near , then .
If , then near 0.
By local boundedness: in some U'(0, δ).
If a limit is positive (or negative), the function has the same sign near the limit point.
If , then :
Take . Then : .
So . Since L > 0: . ∎
If near and limit exists, then .
f(x) < g(x) does NOT imply lim f < lim g.
Example: 0 < x² for x ≠ 0, but lim 0 = lim x² = 0.
If near and both limits exist, then .
If , show f(x) > 3 near x = 2.
Let g(x) = f(x) - 3. Then lim g = 5 - 3 = 2 > 0.
By sign preservation: g(x) > 0 near 2, so f(x) > 3. ∎
If and , then:
Given ε > 0, find δ₁, δ₂ such that |f(x) - a| < ε/2 and |g(x) - b| < ε/2.
For δ = min(δ₁, δ₂):
Find .
Find .
Factor:
Find .
For polynomial P(x): .
For R(x) = P(x)/Q(x) with Q(c) ≠ 0: .
The Squeeze Theorem is powerful for limits involving oscillating functions or products.
If near and , then:
Given ε > 0, find δ such that L - ε < f(x) and h(x) < L + ε near x₀.
Since f ≤ g ≤ h:
So |g(x) - L| < ε. ∎
Find .
Since :
Both bounds → 0. By Squeeze: limit = 0.
Find .
Find .
The Cauchy criterion characterizes limit existence: values get arbitrarily close to EACH OTHER near the limit point.
exists iff:
If lim f = L, find δ: |f(x) - L| < ε/2 for x ∈ U'(x₀, δ).
Then |f(x') - f(x'')| ≤ |f(x') - L| + |f(x'') - L| < ε. ∎
exists iff:
Show does not exist.
Take ε₀ = 1. For any M, let x' = 2πk, x'' = π + 2πk (k large).
Cauchy fails, so limit DNE. ∎
Show exists.
For x', x'' > M = 2/ε:
If and f continuous at L, then .
Find .
Since sin x → 0 and eˣ is continuous: .
If , then .
, so limit = 0.
Find .
Find .
Use equivalences: 1 - cos x ~ x²/2, sin x ~ x
Find .
Solution: Factor using difference of cubes:
Find .
Solution: Let u = ∛x, so x = u³ and x → 8 means u → 2.
As u → 2:
Find .
Solution: Multiply by conjugate:
As x → 0:
Find .
Solution: Use standard limits:
As x → 0: both factors → 1, so limit = 1.
Find .
Solution: Let t = x - 1, so x = 1 + t and x → 1 means t → 0:
Find .
As x → 0: sin x → 0, so tan(sin x)/(sin x) → 1. Also sin x/x → 1.
Limit = 1 × 1 = 1.
Find .
Solution: This is 1^∞ form. Rewrite:
As x → 0, 2x → 0, so (1 + 2x)^{1/(2x)} → e.
Limit = e² .
Find .
Use sin x ~ x, 1 - cos x ~ x²/2, cos x → 1:
Find .
Solution: This is 1^∞ form. Write:
As x → +∞: (x-1)/2 → ∞, so inner part → e.
And 2x/(x-1) → 2. So limit = e².
Find .
Solution: Both sin and cos are bounded by [-1, 1]:
By squeeze: limit = 0.
Let lim f = a and lim g = b. We show lim fg = ab.
Write:
By local boundedness, |f(x)| ≤ M near x₀.
Given ε > 0, choose δ such that:
First prove: if lim g = b ≠ 0, then lim(1/g) = 1/b.
By sign preservation, |g(x)| > |b|/2 near x₀.
Given ε, choose δ: |g(x) - b| < εb²/2. Then |1/g - 1/b| < ε.
Then lim(f/g) = lim(f · 1/g) = (lim f)(lim 1/g) = a · (1/b) = a/b. ∎
Wrong: lim_{x→1} (x²-1)/(x-1) = (1-1)/(1-1) = 0/0 = ???
Correct: Factor first: (x²-1)/(x-1) = x+1 → 2.
Wrong: Since x² > 0 for x ≠ 0, lim x² > 0.
Correct: We can only say lim x² ≥ 0. In fact, lim_{x→0} x² = 0.
Wrong: -1 ≤ sin(1/x) ≤ 1, so -1 ≤ x sin(1/x) ≤ 1.
Correct: Multiply by |x|: -|x| ≤ x sin(1/x) ≤ |x|.
Wrong: lim_{x→1} f(x) = 2 implies f is bounded on ℝ.
Correct: f is only guaranteed bounded NEAR x = 1, not everywhere.
Wrong: lim f(g(x)) = f(lim g(x)) always.
Correct: Only when f is continuous at lim g(x).
The limit properties we've studied are essential for understanding continuity (Chapter 3.3) and its consequences.
Show that f(x) = x² sin(1/x) for x ≠ 0, f(0) = 0 is continuous at 0.
Solution: By squeeze, lim_{x→0} x² sin(1/x) = 0.
Since f(0) = 0 = lim_{x→0} f(x), f is continuous at 0. ∎
If limit exists, it's unique
Limit exists ⟹ |f| ≤ M near x₀
L > 0 ⟹ f(x) > 0 near x₀
f ≤ g ⟹ lim f ≤ lim g
Find .
Solution: Recall x - 1 < ⌊x⌋ ≤ x. Dividing by x (x > 0):
As x → +∞: 1 - 1/x → 1. By Squeeze: limit = 1.
Find .
Solution: Write as product of standard limits:
As x → 0: each sin(kx)/(kx) → 1.
Show for any n ∈ ℕ.
Proof: For x > 2n, we have eˣ > x^{2n}/(2n)! (from Taylor series).
By Squeeze: limit = 0. Exponentials dominate polynomials. ∎
Find .
Solution: |sin(1/x)| ≤ 1 and |cos(1/x²)| ≤ 1, so:
By squeeze: limit = 0.
Find .
Solution: Factor both:
At x = 1:
Find .
Method 1: Factor using a⁵ - b⁵ = (a-b)(a⁴ + a³b + a²b² + ab³ + b⁴):
At x = 2: 16 + 16 + 16 + 16 + 16 = 80.
Method 2: This is f'(2) where f(x) = x⁵ and derivative is 5x⁴ = 5(16) = 80.
Find .
Solution: Rationalize:
Find .
Solution: Factor 1 - cos³x = (1 - cos x)(1 + cos x + cos²x):
Use 1 - cos x ~ x²/2, sin x ~ x, cos x → 1:
Prove does not exist.
Proof: Take ε₀ = 1. For any M > 0, choose k such that √(2πk) > M.
Let x' = √(2πk) and x'' = √(2πk + π/2).
Then (x')² = 2πk and (x'')² = 2πk + π/2.
Cauchy criterion fails, so limit DNE. ∎
Prove exists using Cauchy.
Proof: For x', x'' > M:
Choose M = 2/ε. Then difference < ε. Cauchy is satisfied. ∎
Cauchy criterion is useful when:
For existence, Cauchy doesn't tell you WHAT the limit is—just that it exists.
Uniqueness is derived directly from ε-δ by contradiction
Take ε = 1 in the definition
Take ε = |L|/2 when L ≠ 0
Bounded factor controls the product
Ensures denominator bounded away from 0
Squeeze is a special case of order preservation
Key Formulas:
a² - b² = (a-b)(a+b)
a³ - b³ = (a-b)(a² + ab + b²)
a³ + b³ = (a+b)(a² - ab + b²)
aⁿ - bⁿ = (a-b)(aⁿ⁻¹ + aⁿ⁻²b + ... + bⁿ⁻¹)
Multiply by conjugate when you see √ in numerator or denominator:
This eliminates radicals by creating a difference of squares.
Replace with standard forms when approaching indeterminate:
As x → 0, these are equivalent (use in products only!):
sin x ~ x
tan x ~ x
arcsin x ~ x
arctan x ~ x
1 - cos x ~ x²/2
eˣ - 1 ~ x
ln(1+x) ~ x
(1+x)ᵅ - 1 ~ αx
When you see (1 + something)^{big}, use:
Then use ln(1 + u) ~ u when u → 0:
The limit becomes e^{lim g(x)f(x)}.
Find .
Solution: As x → 0, sin x/x → 1, so this is 1^∞ form.
Write as exponential:
Need to find:
Using Taylor: sin x/x = 1 - x²/6 + O(x⁴), so ln(sin x/x) ~ -x²/6.
Final answer:
Find .
Solution: Taylor: arctan x = x - x³/3 + O(x⁵)
Find .
As x → 0: (eˣ - 1)/x → 1 and ln(1+x)/x → 1.
Limit = 1 · 1 = 1.
Find .
Solution: For x < 0, √(x²) = |x| = -x.
Rationalize: multiply by (1 + √(1+1/x))/(1 + √(1+1/x)):
Find .
Solution: First find lim xˣ as x → 0⁺.
So x^{xˣ} → x¹ = x → 0 as x → 0⁺.
But wait—more carefully: x^{xˣ} = e^{xˣ ln x}.
As x → 0⁺: xˣ → 1 and ln x → -∞, but the product: xˣ · ln x → 1 · (-∞) = -∞.
So x^{xˣ} → e^{-∞} = 0.
Prove uniqueness using ε-δ:
If lim f(x) = L₁ and lim f(x) = L₂, prove L₁ = L₂.
Prove local boundedness:
If lim f(x) = L, prove |f(x)| ≤ |L| + 1 near x₀.
Prove the product rule:
If lim f = a and lim g = b, prove lim(fg) = ab.
Use Cauchy criterion:
Prove lim_{x→∞} sin(x²) does not exist.
When evaluating limits, clever substitutions can simplify problems dramatically:
Many limits reduce to standard forms in disguise:
→ (3/5)·1 = 3/5
→ e⁶
→ 5/3
If direct methods fail, consider:
Many of these limit techniques appear in derivatives:
Understanding limits is essential for understanding calculus!
| Form | Strategy |
|---|---|
| 0/0 | Factor, rationalize, L'Hôpital, Taylor |
| ∞/∞ | Divide by highest power, L'Hôpital |
| 0 · ∞ | Rewrite as 0/0 or ∞/∞ |
| ∞ - ∞ | Combine fractions, factor, rationalize |
| 1^∞ | Use e^{g·ln(1+f)} with ln(1+u) ~ u |
| 0⁰ | Use e^{g·ln f} |
| ∞⁰ | Use e^{g·ln f} |
| Property | Statement | Use |
|---|---|---|
| Uniqueness | If limit exists, it's unique | Foundation |
| Local Boundedness | Limit ⟹ bounded nearby | Product proofs |
| Sign Preservation | L > 0 ⟹ f > 0 nearby | Quotient proofs |
| Order Preservation | f ≤ g ⟹ lim f ≤ lim g | Squeeze theorem |
| Squeeze | f ≤ g ≤ h, lim f = lim h = L ⟹ lim g = L | Oscillating limits |
| Cauchy | |f(x') - f(x'')| < ε near x₀ | Existence proofs |
In CALC-3.3, we'll use these limit properties to study:
Local: bounded near x₀. Global: bounded on entire domain. Limit implies local only.
When denominator limit is 0. Use L'Hôpital or other techniques then.
No! f < g only implies lim f ≤ lim g, not strict inequality.