Intuition The one core idea
A bead slides down a frictionless wire from a high point to a lower point, and we want the wire's shape that makes the trip take the least time. To even ask "which shape is fastest?", we must first learn to add up tiny bits of time along a curve — so every symbol below is a brick in that one sentence.
This page assumes you have seen nothing . We will name every letter, arrow, and squiggle the parent note uses, draw the picture it stands for, and say why the problem needs it. Read top to bottom — each idea leans only on the ones above it.
Definition A point and its coordinates
A point is one exact spot on the page. We pin it down with two numbers ( x , y ) : how far across (x ) and how far down (y ). The pair is called its coordinates .
Look at the first figure. Notice the surprise: the y -axis points downward , not up.
Common mistake Which way does
y point?
Why it feels wrong: in school y always goes up. Fix: here we measure y downward on purpose, so that "deeper" means "bigger y ". This makes the speed formula come out clean (you'll see why in §7). Always state the convention before you start.
A — the start point , placed at the top, at ( 0 , 0 ) .
B — the end point , lower and off to the side.
x 1 — the horizontal distance across to B .
Definition Orientation convention: we move left-to-right
We always travel from A (at x = 0 ) toward B (at x = x 1 ) with x increasing . So each tiny across-step d x is positive , d x > 0 . Fix this now — §4 needs it.
Why the topic needs this: the whole question is "join A to B by a curve" — you cannot join two points without first naming them.
A function y ( x ) is a rule: give it an across-value x , it returns one down-value y . Its picture is a curve — the wire itself.
is the wire
As you walk left-to-right (increasing x ), the wire has exactly one height at each spot. That "one height per x " is precisely what makes it a function. Different wire shapes = different functions y ( x ) .
Definition The wire must stay below the start:
y ( x ) > 0
Because y is measured downward from A , every point of the wire below A has y > 0 . We require y ( x ) > 0 for all x after the start (the wire never climbs back to or above the starting height). This is not fussiness: §7's speed formula contains 2 g y , which is only a real number when y > 0 . The single exception is the start point itself, y ( 0 ) = 0 — handled carefully in §8.
Why the topic needs this: the brachistochrone question is "which y ( x ) ?". The unknown here is not a number — it is an entire curve. Hold that thought; it is the whole reason a new branch of maths (Calculus of Variations ) is needed.
y ′ = d x d y
Zoom into one point of the curve until it looks straight. The slope y ′ is how much it drops (d y ) for a tiny step across (d x ). Written y ′ = d x d y , read "dee-y by dee-x".
d x — an infinitesimally small, positive step across (think "a step so small it's basically a point, but not zero"). By §1 we always have d x > 0 .
d y — the matching tiny drop.
y ′ big ⇒ steep wire; y ′ small ⇒ flat wire; y ′ = 0 ⇒ momentarily level.
Intuition Why we need slope at all
Speed and distance both depend on how tilted the wire is at each spot. The steep top of the winning curve is exactly where y ′ is huge — the slope is how the maths "feels" the shape locally.
Why the "prime" notation? y ′ is short for "the derivative of y " — the tool from calculus that measures instantaneous steepness. We use it because the wire's tilt changes continuously; a single number can't describe it, but a function of x can.
WHAT we did: treated a tiny curve-piece as the hypotenuse of a right triangle with legs d x (across) and d y (down).
WHY: on a scale small enough, any smooth curve looks straight, so Pythagoras applies.
WHAT IT LOOKS LIKE: the little triangle in the figure below.
The second form comes from factoring out d x 2 :
d x 2 + d y 2 = d x 2 ( 1 + d x 2 d y 2 ) = d x 2 1 + y ′2 .
Now d x 2 = ∣ d x ∣ in general — but by our orientation convention from §1 we always move with d x > 0 , so ∣ d x ∣ = d x and
d x 2 + d y 2 = 1 + y ′2 d x .
d x 2 is ∣ d x ∣ , not automatically d x
Why it feels right: "square then square-root cancels". Fix: square returns the positive value, so d x 2 = ∣ d x ∣ . We may drop the bars only because we fixed d x > 0 in §1 (left-to-right travel). Without that convention the sign is genuinely ambiguous.
Common mistake Why not just use
d x for length?
Why it feels right: d x is the horizontal progress. Fix: the bead travels along the slanted wire, not the floor. A steep piece covers little x but lots of actual length — that extra is the 1 + y ′2 factor.
Why the topic needs this: time = length ÷ speed, so we must measure the wire's real length, slant included.
∫ A B ( tiny bit ) means: add up every tiny bit, from A to B , as the bits shrink to zero size. The stretched "S" stands for S um.
Intuition Picture of an integral
Chop the journey into thousands of slivers. Each sliver contributes a tiny time. Stack all those tiny times end to end — the total height of the stack is the integral. As slivers get thinner, the stack's total settles on the exact answer.
∫ 0 x 1 — sum from the start (x = 0 ) to the end (x = x 1 ).
The thing after the ∫ (the integrand ) is "one tiny bit's worth".
Why the topic needs this: total time is built from countless tiny times d t (defined next); only summation (∫ ) assembles them into a single number.
Definition Speed and gravity
Speed v is how fast the bead moves along the wire (length per second). g is the constant pull of gravity that speeds a falling thing up (about 9.8 metres-per-second, every second).
Definition The infinitesimal time
d t
d t is the tiny bit of time the bead spends crossing one tiny arc d s . Since time = distance ÷ speed, for one sliver
d t = v d s .
This is the "one tiny bit's worth" that §5's integral adds up.
v small = crawling; v large = zooming.
The deeper the bead has fallen, the more g has sped it up, so the bigger v becomes.
Why the topic needs this: d t = d s / v is the atom of the whole problem. A fast bead clears the same length in less time — so where the bead is fast matters enormously.
First, meet the mass symbol before we use it.
m
m is the mass of the bead — how much "stuff" it contains (heavier = bigger m ). It appears in both energy terms below and, pleasingly, cancels out.
WHAT we did: set the energy of motion equal to the height-energy given up.
WHY: on a frictionless wire, nothing wastes energy, so "motion-energy gained = fall-energy lost" (Conservation of Energy ).
WHAT IT LOOKS LIKE: deeper (bigger y ) ⇒ faster (bigger v ). The two terms:
2 1 m v 2 — kinetic energy , the energy of moving.
m g y — potential energy released by dropping a height y .
Intuition Why there is no minus sign
Potential energy near the ground is (a constant) − m g h when h points up . But we chose y to point down , so h = − y and the release of energy as the bead descends by y is exactly + m g y — a positive amount gained by the motion. The downward-y choice is precisely what turns the usual minus into a clean plus here. If you had kept y upward, you'd write 2 1 m v 2 = − m g y and juggle signs.
Divide by m and both mass terms vanish — the fastest shape does not depend on the bead's weight. Rearranging gives v = 2 g y .
Intuition Now the downward-
y and y > 0 rules pay off
Because y increases as the bead descends, 2 g y is positive and the square root is real — this is exactly why §2 demanded y > 0 along the wire. Had we pointed y up, we'd be taking negative near the start — a mess.
The square root answers "what number, times itself, gives 2 g y ?" — it undoes the squaring in v 2 , letting us isolate v .
Read it piece by piece with everything above:
top 1 + y ′2 d x = tiny length d s (§4),
bottom 2 g y = speed v (§7),
their ratio = tiny time d t = d s / v (§6),
the ∫ = add up all tiny times (§5).
A functional eats a whole function and returns one number . Here you feed in a wire-shape y ( x ) and get out its total travel time. The square brackets T [ y ] (not round) signal "this depends on the entire function y , not on a single value".
Common mistake "At the start
v = 0 , so d t = d s / v blows up — the time is infinite!"
Why it feels right: at A we have y ( 0 ) = 0 , so v = 2 g y = 0 , and dividing by zero looks fatal. Fix: the bead does start at rest, but it does not linger there — it immediately gathers speed. Near the start the integrand behaves like 1/ y , and the integral of 1/ y is finite (its antiderivative 2 y stays bounded). This is an integrable singularity: infinitely tall for an instant, but enclosing a finite area. So T [ y ] is a perfectly finite number even though the very first sliver's d t formula misbehaves. The winning cycloid, which starts nearly vertical, is exactly the shape that makes the bead pick up speed fastest and tames this corner best.
Why the topic needs this idea: ordinary minimization tweaks a number to shrink something. Here we must tweak a whole curve . That leap is exactly what Calculus of Variations and its master rule, the Euler-Lagrange equation , are built to handle.
You don't solve the ODE on this page, but these symbols appear in the parent — meet them now so they aren't strangers:
θ (theta) — an angle , used as a dial that traces the winning curve (Cycloid ) point by point.
r — the radius of the imaginary rolling circle that draws that cycloid.
C , k — plain constants : fixed numbers pinned down later by forcing the curve through B .
F ( y , y ′ ) — the integrand (the stuff inside ∫ ), also called the Lagrangian (Lagrangian Mechanics ).
∂ (partial-dee) — like a derivative, but wiggle one variable while freezing the others.
∂ F / ∂ y
"How much does F change if I nudge only y and hold y ′ still?" The curly ∂ is the flag for "one variable at a time".
Points and coordinates x y
Function y of x is the wire shape
Arc length ds by Pythagoras
Height y downward positive
Energy conservation gives v
Integral adds tiny times dt
Brachistochrone is a cycloid
Each arrow says "you need the left idea before the right one makes sense". Nothing on the right can be understood by skipping a box on its left.
Cover the right side and test yourself.
What does ( x , y ) mean, and which way does y point here? A point's across-and-down position; y points downward so deeper = bigger y .
Which direction do we travel, and what does that fix about d x ? Left-to-right from A to B , so d x > 0 always.
Why must y ( x ) > 0 along the wire? So that
2 g y (the speed) is a real number;
y = 0 only at the start.
What is a function y ( x ) in this problem? A rule giving one wire-height per across-position — i.e. the shape of the wire.
What does the slope y ′ = d y / d x measure? How steep the wire is at a single point (drop per tiny step across).
Write the tiny arc length and justify dropping the absolute value. d s = 1 + y ′2 d x ;
d x 2 = ∣ d x ∣ = d x because
d x > 0 .
What is d t , and how is it built? The tiny time to cross one arc: d t = d s / v .
What does the symbol ∫ 0 x 1 tell you to do? Add up every tiny piece from x = 0 to x = x 1 as the pieces shrink to zero.
Where does v = 2 g y come from, and what is m ? Energy conservation 2 1 m v 2 = m g y ; m is the bead's mass, which cancels.
Why is there no minus sign in 2 1 m v 2 = m g y ? Because y points down, so the fall-energy released is + m g y .
At the start v = 0 ; why is the total time still finite? The integrand's
1/ y blow-up is integrable (antiderivative
2 y is bounded).
What is a functional, and why T [ y ] not T ( y ) ? A rule turning a whole function into one number; the brackets flag "depends on the entire curve".