This page is a drill of cases . The parent note main topic built the formulas; here we hit every corner where a sign flips, a point degenerates, or an exam adds a twist. First we lay out the full grid of scenarios, then we work one example per grid-cell so you never meet a situation you haven't already seen.
Everything below uses only the tools already earned in the parent note. As a one-line reminder before we start:
Here c is a fixed positive number (the "size" of the hyperbola). The curve has two vertices, at ( c , c ) and at ( − c , − c ) , both lying on the line y = x . And t is just a label for a point — one number that picks out where you are on the curve.
Every problem this topic throws is one of these cells. The last column names the example that covers it.
#
Case class
What is special about it
Covered by
A
t > 0 — Quadrant I branch
both c t > 0 and c / t > 0
Ex 1
B
t < 0 — Quadrant III branch
both coordinates negative
Ex 2
C
Same-sign chord (t 1 , t 2 > 0 )
chord lying on one branch, slope sign check
Ex 3
D
Mixed-sign chord (t 1 > 0 , t 2 < 0 )
chord crossing between the two branches
Ex 4
E
Intersection of two tangents
symmetric-sum formula, sign bookkeeping
Ex 5
F
Degenerate: t 1 = t 2 (chord → tangent)
limiting case, self-check
Ex 6
G
Limiting: t → 0 and t → ∞
point escaping to an asymptote
Ex 7
H
Normal meeting the curve again
cubic in t , third point
Ex 8
I
Real-world word problem
Boyle's law P V = const
Ex 9
J
Exam twist: locus of midpoints
eliminate the parameter
Ex 10
Every example carries a figure so the geometry is always visible.
Worked example Tangent and the two axis-intercepts at
t = 3 on x y = 4
For x y = 4 take t = 3 . Find the tangent line, and where it cuts the x -axis and y -axis. Show the triangle it makes with the axes.
Forecast: Guess before computing — will the midpoint of that axis-cut segment land on the curve, or somewhere else?
Read off c . c 2 = 4 ⇒ c = 2 . The point is ( c t , t c ) = ( 6 , 3 2 ) .
Why this step? Every formula is written in c and t ; we must know c before substituting.
Write the tangent with t x + t y = 2 c : 3 x + 3 y = 4 .
Why this step? This is the earned tangent form — no re-differentiating.
x -intercept (set y = 0 ): 3 x = 4 ⇒ x = 12 , giving ( 12 , 0 ) .
y -intercept (set x = 0 ): 3 y = 4 ⇒ y = 3 4 , giving ( 0 , 3 4 ) .
Why this step? Intercepts are the natural "corners" of the tangent-with-axes triangle.
Midpoint of intercepts: ( 2 12 + 0 , 2 0 + 4/3 ) = ( 6 , 3 2 ) .
Why this step? Compare it to the point of tangency from step 1.
Answer: Tangent x + 9 y = 12 ; intercepts ( 12 , 0 ) and ( 0 , 3 4 ) ; their midpoint is ( 6 , 3 2 ) — exactly the point of contact . (So the forecast answer is: yes, the point of contact bisects the tangent's intercept segment — a signature property.)
Verify: Point on line? 6 + 9 ⋅ 3 2 = 6 + 6 = 12 ✓. Point on curve? 6 ⋅ 3 2 = 4 = c 2 ✓.
Worked example Tangent at
t = − 2 on x y = 4 — watch the signs
Same curve x y = 4 , but now t = − 2 . Find the point and tangent, and confirm the slope is still negative.
Forecast: The point is in Quadrant III (both coords negative). Will the tangent's slope be positive or negative there?
Point: ( c t , t c ) = ( 2 ( − 2 ) , − 2 2 ) = ( − 4 , − 1 ) .
Why this step? With t < 0 both c t and c / t come out negative — that is precisely the third quadrant branch.
Slope: − t 2 1 = − ( − 2 ) 2 1 = − 4 1 .
Why this step? t 2 > 0 always, so the slope is negative on both branches — a common exam trap (mistake (b) in the parent note).
Tangent: t x + t y = 2 c ⇒ − 2 x + ( − 2 ) y = 4 , multiply by − 2 : x + 4 y = − 8 .
Why this step? Keep the sign of t inside the formula — do not "make it positive."
Answer: Point ( − 4 , − 1 ) , slope − 4 1 , tangent x + 4 y = − 8 .
Verify: On curve: ( − 4 ) ( − 1 ) = 4 = c 2 ✓. On line: − 4 + 4 ( − 1 ) = − 8 ✓. Slope from line: y = − 4 1 x − 2 , slope − 4 1 ✓.
Worked example Chord joining
t 1 = 1 and t 2 = 2 on x y = 4 (both in Q I)
Find the chord line connecting two points on the same first-quadrant branch, and note its slope sign.
Forecast: Both points sit on the descending Q I branch. Positive or negative slope?
Endpoints: t 1 = 1 ⇒ ( 2 , 2 ) ; t 2 = 2 ⇒ ( 4 , 1 ) .
Why this step? Both coordinates positive — a chord living entirely on one branch.
Chord formula x + t 1 t 2 y = c ( t 1 + t 2 ) with t 1 t 2 = ( 1 ) ( 2 ) = 2 and t 1 + t 2 = 3 :
x + 2 y = 2 ( 3 ) = 6.
Why this step? The symmetric chord form again, now with a positive product.
Slope: y = − 2 1 x + 3 , slope − 2 1 .
Why this step? − t 1 t 2 1 = − 2 1 — negative because t 1 t 2 > 0 . On a single branch the curve only ever falls, so a same-sign chord always has negative slope.
Answer: Chord x + 2 y = 6 , slope − 2 1 .
Verify: ( 2 , 2 ) : 2 + 4 = 6 ✓. ( 4 , 1 ) : 4 + 2 = 6 ✓.
Worked example Chord joining
t 1 = 1 (Q I) and t 2 = − 2 (Q III) on x y = 4
Find the chord line connecting a point on the first-quadrant branch to one on the third-quadrant branch.
Forecast: A line joining the two opposite branches — do you expect a positive or negative slope? (Trust your picture, not a formula, first.)
Endpoints: t 1 = 1 ⇒ ( 2 , 2 ) ; t 2 = − 2 ⇒ ( − 4 , − 1 ) .
Why this step? We need the two actual points to sanity-check the chord equation.
Chord formula x + t 1 t 2 y = c ( t 1 + t 2 ) with t 1 t 2 = ( 1 ) ( − 2 ) = − 2 and t 1 + t 2 = − 1 :
x − 2 y = 2 ( − 1 ) = − 2.
Why this step? The symmetric chord form saves the two-point-slope grind.
Slope from it: y = 2 1 x + 1 , slope + 2 1 .
Why this step? Here − t 1 t 2 1 = − − 2 1 = + 2 1 — a positive slope because t 1 t 2 < 0 . Matches the picture: a line climbing from Q III up to Q I.
Answer: Chord x − 2 y = − 2 , slope + 2 1 .
Verify: ( 2 , 2 ) : 2 − 4 = − 2 ✓. ( − 4 , − 1 ) : − 4 + 2 = − 2 ✓.
Worked example Where do the tangents at
t 1 = 1 and t 2 = 4 on x y = 4 meet?
Find the meeting point of the two tangent lines.
Forecast: Both t 's are positive, so both contact points are in Q I. Will their tangents cross inside Q I too?
Use the recalled intersection formula ( t 1 + t 2 2 c t 1 t 2 , t 1 + t 2 2 c ) (sixth formula above) with c = 2 , t 1 t 2 = 4 , t 1 + t 2 = 5 :
( 5 2 ⋅ 2 ⋅ 4 , 5 2 ⋅ 2 ) = ( 5 16 , 5 4 ) .
Why this step? This is the ready-made result; deriving from scratch every time wastes exam minutes.
Cross-check by intersecting the lines directly. Tangents: 1 x + 1 ⋅ y = 4 , i.e. x + y = 4 ; and 4 x + 4 y = 4 , i.e. x + 16 y = 16 . Subtract: 15 y = 12 ⇒ y = 5 4 , then x = 4 − 5 4 = 5 16 .
Why this step? Two independent routes agreeing is our proof.
Answer: They meet at ( 5 16 , 5 4 ) — inside Q I, as forecast.
Verify: 5 16 + 5 4 = 5 20 = 4 ✓ (on first tangent); 5 16 + 16 ⋅ 5 4 = 5 16 + 5 64 = 5 80 = 16 ✓ (on second).
Worked example Force the chord through
t 1 = t 2 = 3 and confirm it is the tangent
Take the chord formula and let both parameters merge to t = 3 on x y = 4 . Show it collapses to the tangent at t = 3 .
Forecast: A chord needs two points. If they merge into one, geometrically the secant line should rotate into the tangent. Do the algebra agree?
Chord x + t 1 t 2 y = c ( t 1 + t 2 ) with t 1 = t 2 = 3 , c = 2 : t 1 t 2 = 9 , t 1 + t 2 = 6 , so
x + 9 y = 12.
Why this step? Setting the two labels equal is exactly the limiting operation "second point slides onto the first."
Tangent at t = 3 (from Ex 1): 3 x + 3 y = 4 , multiply by 3 : x + 9 y = 12 .
Why this step? If the chord limit is meaningful it MUST reproduce the tangent — and it does, identically.
Answer: Chord→ Tangent gives the same line x + 9 y = 12 . The degenerate case is consistent.
Verify: Point of contact ( 6 , 3 2 ) on x + 9 y = 12 : 6 + 9 ⋅ 3 2 = 12 ✓.
Worked example What happens to the point
( c t , c / t ) at the extremes of t ? (x y = 4 )
Describe the point as t → 0 + , t → + ∞ , and the negative mirrors, and connect this to the asymptotes.
Forecast: Both asymptotes are the coordinate axes x = 0 and y = 0 . Which extreme of t hugs which axis?
As t → + ∞ : c t = 2 t → + ∞ but t c = t 2 → 0 + . The point races rightward along y = 0 + — it approaches the x -axis (an asymptote) from above.
Why this step? This is why y = 0 is an asymptote: the curve gets arbitrarily close but y never equals 0 (would need x = ∞ ).
As t → 0 + : c t = 2 t → 0 + but t c = t 2 → + ∞ . The point shoots up along x = 0 + — approaching the y -axis asymptote.
Why this step? Symmetric reason: x = 0 is the other asymptote.
Negative mirror (t → 0 − , t → − ∞ ): same behaviour reflected through the origin, hugging the axes from the Q III side.
Why this step? Covers the whole matrix cell — no extreme left unshown.
Answer: t → ± ∞ presses the curve onto the x -axis; t → 0 ± presses it onto the y -axis. The two coordinate axes are the asymptotes.
Verify (numeric squeeze): at t = 1000 : point ( 2000 , 0.002 ) , product = 4 ✓ and y tiny. At t = 0.001 : point ( 0.002 , 2000 ) , product = 4 ✓ and x tiny.
Worked example The normal at
t = 1 on x y = 4 meets the curve again at parameter t ′ — find it
A normal is drawn at t = 1 . It cuts the hyperbola a second time. Find that other parameter and point.
Forecast: A line through a hyperbola meets it in (at most) two points, so exactly one other t ′ should appear. Guess: bigger or smaller than 1 ?
Normal at t = 1 : x t 3 − y t = c ( t 4 − 1 ) with t = 1 , c = 2 gives x ( 1 ) − y ( 1 ) = 2 ( 1 − 1 ) = 0 , i.e. y = x .
Why this step? At the vertex t = 1 the point is ( 2 , 2 ) on the axis y = x , so the normal is the axis itself — a clean, checkable case.
Substitute the parametric point ( 2 t ′ , 2/ t ′ ) into y = x : t ′ 2 = 2 t ′ ⇒ t ′2 = 1 ⇒ t ′ = ± 1 .
Why this step? Any second intersection must satisfy both the line and the parametrization.
Discard t ′ = 1 (that is the starting point). The other root is t ′ = − 1 , giving ( − 2 , − 2 ) .
Why this step? The starting parameter is always one root of the intersection equation; the new information is the second root.
Answer: The normal at t = 1 (the line y = x ) meets the curve again at t ′ = − 1 , point ( − 2 , − 2 ) — the opposite vertex.
Verify: ( − 2 , − 2 ) on curve: ( − 2 ) ( − 2 ) = 4 ✓; on y = x : − 2 = − 2 ✓.
General fact (worth knowing): the normal at t meets the curve again at t ′ = − t 3 1 ; at t = 1 that is − 1 , matching us exactly.
Worked example Gas at constant temperature:
P V = 48 (units: kPa·L)
A fixed amount of gas at constant temperature obeys Boyle's Law : pressure × volume is constant, P V = 48 . (a) If V = 6 L , find P . (b) The engineer needs the "rate" d V d P at that state. (c) Interpret the constant-area property physically.
Forecast: P V = const is literally x y = c 2 with x = V , y = P , c 2 = 48 . So d P / d V should be negative (squeeze the gas, pressure rises).
Find P : P = V 48 = 6 48 = 8 kPa .
Why this step? This is just the reciprocal-function value y = c 2 / x (see Reciprocal function y=1/x ).
Rate: with P V = 48 , implicit differentiation gives d V d P = − V P = − 6 8 = − 3 4 kPa/L .
Why this step? Same d x d y = − x y as the curve — the physics inherits the geometry. The minus sign means compressing (smaller V ) raises P .
Constant-area meaning: the rectangle from the state point ( V , P ) back to the axes has area V ⋅ P = 48 — the "same-area rule" is exactly Boyle's constant.
Why this step? It ties the abstract rectangle property to a measured physical invariant.
Answer: P = 8 kPa , d V d P = − 3 4 kPa/L , and the state-rectangle area is the fixed 48 .
Verify: 6 × 8 = 48 ✓ (units kPa·L consistent). Rate sign negative as expected ✓.
Worked example Midpoint of a chord of
x y = 4 that touches parameters t 1 , t 2 with t 1 t 2 = 1 — find its locus
As t 1 , t 2 vary but always with product t 1 t 2 = 1 , the chord's midpoint traces some curve. Find its equation.
Forecast: A constraint like t 1 t 2 = 1 usually forces the midpoint onto a straight line or a scaled hyperbola. Which?
Midpoint coordinates:
x m = 2 c t 1 + c t 2 = 2 c ( t 1 + t 2 ) , y m = 2 c / t 1 + c / t 2 = 2 t 1 t 2 c ( t 1 + t 2 ) .
Why this step? Average the two parametric points; write everything in the symmetric quantities t 1 + t 2 and t 1 t 2 .
Apply the constraint t 1 t 2 = 1 : then y m = 2 ⋅ 1 c ( t 1 + t 2 ) = 2 c ( t 1 + t 2 ) = x m .
Why this step? Substituting the given product collapses y m directly onto x m .
Answer: The midpoint satisfies y m = x m : the locus is the line y = x — the transverse axis of the hyperbola. (So the forecast "straight line" wins.)
Verify (numeric): take t 1 = 2 , t 2 = 2 1 (product 1 ), c = 2 : points ( 4 , 1 ) and ( 1 , 4 ) , midpoint ( 2 5 , 2 5 ) — lies on y = x ✓.
Recall Which cell was hardest — quick self-quiz
Chord joining t 1 > 0 , t 2 < 0 has slope of what sign? ::: Positive, since slope = − t 1 t 2 1 and t 1 t 2 < 0 .
Chord with both t 1 , t 2 > 0 has slope of what sign? ::: Negative, since t 1 t 2 > 0 ; a single branch only falls.
As t → ∞ on x y = c 2 the point hugs which axis? ::: The x -axis (y → 0 ).
Normal at t re-meets the curve at what parameter? ::: t ′ = − t 3 1 .
Setting t 1 = t 2 in the chord formula gives? ::: The tangent at that t .
Boyle's law P V = const is which conic? ::: A rectangular hyperbola x y = c 2 .
Parent: Rectangular hyperbola xy = c² — all formulas drilled here.
Reciprocal function y=1/x — the Boyle's-law example is a scaled version of this graph.
Tangent and Normal to conics — Examples 1, 2, 8 exercise the tangent/normal machinery.
Asymptotes of a hyperbola — Example 7 shows the limiting escape onto them.
Hyperbola standard form · Rotation of axes · Eccentricity — background for why these formulas hold.