Calculus II ep12: L'Hospital's rule (Oct 5, 2026)

Overview

Prof Staecker reviews inverse tangent, inverse sine, and inverse cosine functions before deriving their key differentiation and integration formulas, emphasizing chain rule and u-substitution techniques for expressions such as 1/(1+x²) and 1/√(1−x²). He then introduces L'Hôpital's rule for 0/0 indeterminate limits, applies it to rational, trigonometric, and logarithmic examples, explains why the rule works through local linearization, and highlights when repeated application is invalid.

Key takeaways

The most useful inverse-trigonometric integration formulas introduced are ∫ dx/√(1−x²) = arcsin(x) + C and ∫ dx/(1+x²) = arctan(x) + C.
For scaled inverse-trigonometric integrals, constants must be handled structurally: 4x² becomes (2x)², while 16+x² becomes 16[1+(x/4)²].
L'Hôpital's rule requires checking direct substitution first; it is valid for the 0/0 indeterminate form, not for every quotient limit.
The rational limit limₓ→₁ (x²+2x−3)/(x²−3x+2) equals −4 both by factoring and cancellation and by differentiating numerator and denominator.
The limits sin(x)/x as x approaches 0 and [cos(x)−1]/x as x approaches 0 evaluate to 1 and 0, respectively, after one application of L'Hôpital's rule.
A second application of L'Hôpital's rule is not automatically justified: in [cos(x)−1]/√x, the first derivative creates x^(−1/2), which is undefined at zero, so algebraic rewriting is required before evaluating the limit.

Chapters

0:00 Inverse Tangent Graph, Domain, and Range

Prof Staecker contrasts tangent's vertical asymptotes at x = ±π/2 with arctangent's horizontal asymptotes at y = ±π/2.
The arctangent is formed by reflecting the one-to-one branch of tangent across y = x.
arctan(x) has domain (−∞, ∞) and range (−π/2, π/2); calculator results use radians only when the calculator is in radian mode.

6:00 Derivative of arctan(x) and Nested Chain Rules

The derivative formula is d/dx[arctan(x)] = 1/(1+x²), obtained through implicit differentiation and a right-triangle/Pythagorean relationship.
Unlike arcsin(x) and arccos(x), arctan(x) produces no square root because tangent uses opposite/adjacent rather than the hypotenuse.
For arctan(5x²), the chain rule multiplies 1/(1+(5x²)²) by 10x; nested inputs such as ln(3x+1) require another factor of 3x-dependent differentiation.

10:00 Inverse-Trigonometric Antiderivative Formulas

The derivative of arcsin(x) gives the antiderivative formula ∫ dx/√(1−x²) = arcsin(x) + C.
The corresponding arccos formula is valid but less useful because d/dx[arccos(x)] = −1/√(1−x²).
The second central formula is ∫ dx/(1+x²) = arctan(x) + C; ordinary u-substitution alone cannot discover these inverse-trigonometric forms.

14:00 Definite arctan Integral and the Appearance of π

Evaluating ∫₀¹ dx/(1+x²) gives arctan(1) − arctan(0).
Unit-circle knowledge simplifies arctan(1) to π/4 and arctan(0) to 0.
The final value is π/4, illustrating that π can arise from an integral without an obvious circle in the original integrand.

18:00 Absorbing Constants into the Square for Arcsine Integrals

For an arcsine-type denominator containing 4x², Prof Staecker rewrites 4x² as (2x)² so the standard 1−u² pattern becomes visible.
The substitution u = 2x gives du = 2 dx and dx = 2 du, producing a factor of 1/2 outside the integral.
The result has the form (1/2)arcsin(2x) + C; a coefficient such as 5 would require √5x inside the square.

22:00 Scaling arctan Integrals with a Constant Inside 1 + x²

For an integrand involving 16+x², factoring out 16 converts the denominator into 16[1+(x/4)²].
The substitution u = x/4 gives dx = 4 du, and the constants combine to produce (1/4)arctan(x/4) + C.
The general pattern is that a squared scale factor in the denominator becomes its square root in the arctangent argument and an accompanying coefficient.

26:00 Square-Root Scaling and the Arcsine Example

When a constant appears inside a square root, it must be extracted as a square root; √4 contributes a factor of 2 rather than 4.
Rewriting the expression with (x/2)² allows the substitution u = x/2, followed by dx = 2 du.
After simplifying the constants, the worked antiderivative is 3 arcsin(x/2) + C, with errors commonly arising from mishandling dx or √4.

30:00 Transition from Inverse Trigonometry to L'Hôpital's Rule

Prof Staecker closes the inverse-trigonometric section by noting that arctan integrals occur more frequently than arcsine integrals in typical calculus problems.
L'Hôpital's rule is introduced as a limit technique, not an integration method, despite its placement in a Calculus II chapter.
The rule applies when direct substitution into f(x)/g(x) produces the indeterminate form 0/0.

34:00 L'Hôpital's Rule for 0/0 Limits

If f(a) = 0 and g(a) = 0, then under the rule's conditions, limₓ→ₐ f(x)/g(x) equals limₓ→ₐ f′(x)/g′(x).
Direct substitution must come first; differentiating numerator and denominator arbitrarily is invalid when the original limit is not 0/0.
The rule replaces factoring and cancellation, which are the standard introductory-calculus response to a 0/0 result.

40:00 Rational Example: Factoring versus Differentiating

For limₓ→₁ (x²+2x−3)/(x²−3x+2), substitution gives 0/0.
Factoring yields (x+3)(x−1)/[(x−2)(x−1)], so cancellation gives the value −4 at x = 1.
L'Hôpital's rule reaches the same result from (2x+2)/(2x−3), which evaluates to 4/(−1) = −4.

46:00 Why L'Hôpital's Rule Works: Local Linearization

Near a point where f(a) = g(a) = 0, zooming in makes the graphs of f and g resemble straight lines through the origin.
The ratio f(x)/g(x) becomes a ratio of vertical heights, while f′(x)/g′(x) becomes a ratio of the corresponding slopes.
The rule follows because, sufficiently near a, the ratio of the functions approaches the ratio of their local linear approximations.

53:00 Logarithmic and Trigonometric Limits, Including a Pitfall

For limₓ→₁ ln(x)/(x−1), substitution gives 0/0; differentiating produces (1/x)/1, so the limit is 1.
For limₓ→₀ [cos(x)−1]/x, L'Hôpital's rule gives (−sin x)/1 and therefore 0; for sin(x)/x, it gives cos(x)/1 and therefore 1.
For [cos(x)−1]/√x, the first application produces −sin x divided by (1/2)x^(−1/2), not 0/0 because x^(−1/2) is undefined at x = 0; rewriting gives −2√x sin x, whose limit is 0.
Repeated L'Hôpital differentiation is allowed only after another genuine 0/0 form; a denominator involving 0 raised to a negative power is undefined rather than zero.

Summary, takeaways, and chapters were generated by AI from the video's transcript and may contain errors. The video belongs to its creator, Prof Staecker.