Before you can even read the parent note, you need every squiggle it uses to feel obvious. Below is every symbol and idea, built in the order that each one leans on the previous. Nothing is used before it is drawn.
Read the figure below: the blue line is one such curve. The red dot marks a single input-output pair "(x,y)" — slide the dot along the curve and you read off every value the machine produces. When we "solve an ODE" we are hunting for which blue curve fits a given rule.
Why the topic needs it: the whole chapter is about finding an unknown y(x). Everything else describes how y bends and climbs.
Figure 1 — A function y(x): each input x lands you at one point (x,y) on the curve.
Read the figure below: the orange straight line is a ruler laid so it just kisses the curve at the orange dot (the tangent line). The tilt of that ruler isy′ at that point — steep ruler means large y′.
Why "prime"? The little tick mark ′ is just shorthand for "the slope-function of". It is itself a new function of x: at every x it hands back a slope.
Why the topic needs it: an ODE is a sentence written in slopes. You cannot read it without knowing y′ means "slope".
Read the figure below: the same blue curve is a valley ⌣. Because it cups upward everywhere, its bending y′′ is positive (the green note). The orange tangent's tilt keeps increasing as you slide right — that increase isy′′.
Notation ladder:y(k) just means "differentiate k times". So y(0)=y, y(1)=y′, y(2)=y′′, and y(n) is the general n-th one.
Figure 2 — y′ is the tilt of the tangent ruler; y′′ is how fast that tilt keeps changing.
Read the figure below: three exponentials. The red curve (r>0) climbs faster and faster; the green curve (r<0) fades toward the axis; the dashed gray line (r=0) is the flat constant e0=1. At the orange dot on the red curve, the tangent's steepness equals r times the curve's height — differentiating only rescales, never reshapes.
The role of r:r is the dial that sets the exponential's behaviour. r>0 → grows, r<0 → decays toward zero, r=0 → flat constant e0=1.
Figure 3 — erx for r>0, r<0, r=0: differentiating just multiplies by r.
Now we do the crucial move the parent note relies on: plug the guess in and watch the calculus turn into algebra.
Step 1 — Write the guess and its derivatives. Using the self-copying rule from Section 4,
y=erx,y′=rerx,y′′=r2erx.Why: each derivative of erx is just erx multiplied by another factor of r.
Step 2 — Substitute into ay′′+by′+cy=0.a(r2erx)+b(rerx)+c(erx)=0.Why: the equation demands the mix be zero; we simply replace y,y′,y′′ by their formulas.
Step 3 — Factor out the shared erx.erx(ar2+br+c)=0.Why: every term contains one copy of erx (this is exactly the "same shape" point) — pull it out front.
Step 4 — Divide by erx, which is never zero. Since erx>0 for every real x, the product can only vanish if the bracket vanishes:
ar2+br+c=0Why: the calculus problem has collapsed into a plain quadratic. This is the characteristic (auxiliary) equation. Its roots r are the dial-settings that make the guess actually solve the ODE.
Picture it: imagine sliders a, b, c you set once and lock. If instead they were a(x), b(x) (changing with x), the erx trick from Section 5 would break — the factor-out step would no longer leave a clean polynomial. So this restriction is exactly what makes the chapter's method work.
Why the leading one a=0 matters: if a=0 the equation isn't second order anymore. The root formula r=2a−b±b2−4ac (built in Section 8) divides by a — miss it and every root is wrong.
Why it shows up: when the number-puzzle ar2+br+c=0 has a negative discriminant (Section 9), its roots involve −1=i. They come in conjugate pairsα+iβ and α−iβ (mirror images across the horizontal axis).
What each part will mean later:α becomes the growth/decay rate of the solution's envelope; β becomes the wiggle frequency. Euler's formula eiβx=cosβx+isinβx is the bridge from complex r to real waves — see Euler's Formula and Complex Exponentials.
Why n constants: each differentiation you undo introduces one "integration constant" of freedom, and you fix them later using initial conditions like y(0)=1. The tool that formally checks independence is the Wronskian and Linear Independence of Solutions.