Huygens' Principle & Wavefronts

Hold a page of this text at arm's length and it looks sharp; the light that carries its image travelled in straight lines to your eye. Yet the same light, squeezed through a slit narrower than a hair, fans out into soft bands — it bends. How can one thing march in straight lines through an open room and spread like ripples through a narrow gap? The answer that opens the whole of wave optics is almost embarrassingly simple, and it is nearly 350 years old. In 1678 Christiaan Huygens proposed that we stop tracking rays and start tracking wavefronts — surfaces on which the wave is doing exactly the same thing at the same instant, all in step, all at the same phase. A ray is just the arrow drawn at right angles to a wavefront, pointing where the wave is headed. Huygens' one bold idea was that every point on a wavefront is itself a tiny source: it sends out a little spherical wavelet that crawls forward at the wave speed. A short time later, the new wavefront is simply the surface that touches all those wavelets at once — their common tangent, their envelope. That is the entire engine of this article. From nothing but "each point re-radiates, and we take the envelope," pure geometry hands us the law of reflection (angle in equals angle out) and Snell's law of refraction — no forces, no Maxwell's equations, just wavelets and the ruler. It also tells us the honest, often-forgotten truth about a wave crossing into glass or water: its frequency never changes, but it slows down and its wavelength shrinks. Master the construction here and the double slit, thin films, and diffraction in the rest of this chapter become the same picture, drawn again.

Key takeaways

A wavefront is a surface of constant phase; a ray is the line drawn perpendicular to it, in the direction of travel.
Huygens' principle: every point on a wavefront acts as a source of secondary spherical wavelets that travel at the wave speed $v$ ; the new wavefront a time $t$ later is their common tangent (envelope).
From equal wavelet radii on a flat mirror, geometry gives the law of reflection, angle of incidence $=$ angle of reflection.
From wavelets travelling at different speeds across a boundary, geometry gives Snell's law, $\dfrac{\sin\theta_1}{\sin\theta_2}=\dfrac{v_1}{v_2}=\dfrac{n_2}{n_1}$ .
On refraction the source fixes the frequency, so $f$ is unchanged; the speed becomes $v=c/n$ and the wavelength shrinks to $\lambda_{\text{medium}}=\lambda_0/n$ .
Light "goes straight" only because $\lambda$ is far smaller than everyday openings; shrink the opening toward $\lambda$ and the wavelets no longer cancel sideways — the wave spreads (diffraction).
Key idea: treat every point of a wavefront as a fresh wavelet source and take the envelope, and reflection and refraction fall out of geometry alone — with the frequency fixed and only the speed and wavelength changing across a boundary.

The core insight

Every point on a wavefront re-radiates a secondary wavelet at the local wave speed, and the wave a moment later is the envelope of those wavelets — a construction that yields the law of reflection ( $i=r$ ) and Snell's law ( $\sin\theta_1/\sin\theta_2 = v_1/v_2 = n_2/n_1$ ) from pure geometry, with the frequency unchanged and the wavelength shrinking to $\lambda_0/n$ inside a medium.

Where this lives today

The Huygens construction is not a museum piece; it is how physicists still think about wave propagation before reaching for heavier mathematics. Every optical-design program that traces wavefronts through a camera lens, every ultrasound scanner that reconstructs an image from echoes, and every phased-array radar or 5G antenna that steers a beam by delaying the wavelets from each element is applying "re-radiate and take the envelope" in software. Seismologists trace earthquake wavefronts through the Earth's layers by exactly Huygens' method — a boundary between rock types refracts a seismic wave just as glass refracts light, and reading those bent wavefronts is how we mapped the planet's core. The same principle underlies the beam-forming in a modern ultrasound probe and the "wavefront sensing" that lets giant telescopes cancel atmospheric blur in real time. And the seed idea of this article — that entering a medium shrinks the wavelength to $\lambda_0/n$ while leaving the frequency alone — is the foundation of every interference and thin-film calculation in the chapters that follow.

The physical picture

Drop a pebble in a still pond and circular ripples spread out. Freeze one instant: the crest that is rising everywhere along a certain circle is a wavefront — every point of it is at the same stage of its up-and-down motion, the same phase. Far from the pebble a small patch of that circle looks flat, a plane wavefront; close in it is curved, a spherical wavefront. Light from a distant star arrives as near-perfect plane wavefronts; light from a nearby bulb arrives as expanding spheres. In every case the ray is just the arrow perpendicular to the wavefront, showing the direction of travel. Wavefront and ray are two descriptions of the same wave — one a surface, one a line at right angles to it. Huygens asked a simple question: given a wavefront now, where is the wavefront a short time $t$ later? His answer is the principle that carries his name. Treat every point of the present wavefront as a point source of a secondary spherical wavelet, spreading outward at the wave speed $v$ , so that after time $t$ each wavelet is a little sphere of radius $vt$ . The new wavefront is the surface that just touches — is tangent to — all those wavelets at once. Where the medium is uniform and the old wavefront is a plane, the wavelets all have the same radius $vt$ and their common tangent is a parallel plane a distance $vt$ ahead: the wave marches straight on. The principle looks almost trivial there, but its power shows the instant the wavelets are made to travel unequal distances or at unequal speeds — a mirror or a glass boundary — because then their envelope tilts, and that tilt is reflection and refraction.

Figure

A plane wavefront advancing by Huygens construction — each point on the old front emits a secondary wavelet of radius v t, and the new front is their common tangent

Reflection: equal wavelets, equal angles

Send a plane wavefront at a flat mirror at an angle. The edge that reaches the mirror first starts emitting its secondary wavelet immediately, while the rest of the wavefront is still approaching. By the time the far edge lands on the mirror, the first wavelet has grown to a definite radius, and the reflected wavefront is the tangent drawn from the far edge back to that wavelet. Because every point emits at the same speed $v$ , the wavelet radii and the still-to-travel distances match up exactly, and the geometry forces the reflected wavefront to lean away from the mirror at precisely the angle the incident one leaned toward it. Measured from the normal (the line perpendicular to the mirror), the angle of incidence equals the angle of reflection:

\theta_i = \theta_r \qquad (1)

where: $\theta_i$ is the angle between the incoming ray and the normal, and $\theta_r$ the angle between the outgoing ray and the normal. The physical reading is that a mirror does not "know" about forces or fields in this account — the equal-angle law is nothing more than the statement that all the wavelets crawl at one speed, so a symmetric bounce is the only tangent geometry allows. A ray striking at $30^\circ$ to the normal leaves at $30^\circ$ on the other side.

Refraction: unequal speeds bend the front

Now let the wavefront cross from air into glass. The trick is that the wave travels slower in the denser medium — its speed is set by the refractive index through $v=c/n$ , so in glass with $n=1.5$ the speed drops from $c=3.0\times10^{8}\ \mathrm{m\,s^{-1}}$ to

v = \frac{c}{n} = \frac{3.0\times10^{8}}{1.5} = 2.0\times10^{8}\ \mathrm{m\,s^{-1}} \qquad (2)

where: $c$ is the speed of light in vacuum (we use the high-school value $3.0\times10^{8}\ \mathrm{m\,s^{-1}}$ ), $n$ the refractive index, and $v$ the speed in the medium. Physically the wave has slowed to two-thirds of its vacuum speed. Here is the consequence for the wavefront: the edge that enters the glass first is immediately slowed, so its wavelet only reaches radius $v_2 t$ while the edge still in air races ahead by the larger distance $v_1 t$ . The two edges advance by different amounts, the tangent connecting the new wavelets tilts, and the wavefront — and its ray — bend toward the normal. Working the triangles (done in full below) gives Snell's law:

\frac{\sin\theta_1}{\sin\theta_2} = \frac{v_1}{v_2} = \frac{n_2}{n_1} \qquad (3)

where: $\theta_1$ is the angle of incidence and $\theta_2$ the angle of refraction, both from the normal; $v_1,v_2$ are the wave speeds and $n_1,n_2$ the indices of the two media. The last equality uses $v=c/n$ : the slower side (bigger $n$ ) is the side toward which the ray bends. For air ( $n_1=1$ ) into glass ( $n_2=1.5$ ) at $\theta_1=30^\circ$ , with $\sin 30^\circ = 0.5$ ,

\sin\theta_2 = \frac{n_1}{n_2}\sin\theta_1 = \frac{0.5}{1.5} = 0.3333 \quad\Rightarrow\quad \theta_2 = 19.47^\circ,

so the ray bends from $30^\circ$ down to about $19.5^\circ$ — closer to the normal, exactly as "slower in the denser medium" demands. The ratio Huygens' construction actually delivers is $v_1/v_2 = n_2/n_1 = 1.5$ .

Figure

The refraction construction — while the trailing edge travels v1 t in air, the wavelet from A spreads only v2 t in glass, tilting the refracted wavefront and bending the ray toward the normal

What changes, and what does not, across a boundary

A wave crossing into glass is governed by one rule that students forget under exam pressure: the source sets the frequency, and the boundary cannot change it. The atoms of the far medium are driven by the wave arriving at the interface; they must oscillate at that same rate, so $f$ passes through untouched. For $\lambda_0 = 600\ \mathrm{nm}$ light,

f = \frac{c}{\lambda_0} = \frac{3.0\times10^{8}}{600\times10^{-9}} = 5.0\times10^{14}\ \mathrm{Hz} \qquad (4)

where: $\lambda_0$ is the vacuum wavelength and $f$ the frequency, which stays $5.0\times10^{14}\ \mathrm{Hz}$ in air, glass, or water alike. What does change is the speed, and since $v=f\lambda$ with $f$ fixed, the wavelength must shrink in step with the speed:

\lambda_{\text{medium}} = \frac{v}{f} = \frac{c/n}{f} = \frac{\lambda_0}{n}, \qquad \lambda_{\text{glass}} = \frac{600\ \mathrm{nm}}{1.5} = 400\ \mathrm{nm}.

So in glass the same yellow-orange light has a wavelength of $400\ \mathrm{nm}$ — two-thirds of its vacuum value — while its frequency, and therefore its colour, is unchanged. This one fact ("optical path counts $n$ times the geometric path because $\lambda$ is $n$ times shorter") is the seed of every interference calculation later in the chapter. Finally, Huygens' picture explains why light usually seems to travel in straight lines at all. A wavefront passing a wide opening re-radiates wavelets across the whole gap; sideways, the wavelets from all those points cancel by superposition, and only the straight-ahead envelope survives — rectilinear propagation. Narrow the opening toward the wavelength itself and there are too few wavelet sources to cancel sideways, the cancellation fails, and the wave spreads. That spreading is diffraction, and it is where this chapter is heading.

Derivation AThe law of reflection from Huygens wavelets

The goal is to prove that the angle of incidence equals the angle of reflection using nothing but the wavelet construction on a flat mirror — no assumption about forces, no calculus, only the fact that every point of the wavefront re-radiates at the same speed $v$ and that the reflected wavefront is the common tangent to those wavelets. We follow one plane wavefront from the moment its leading edge touches the mirror to the moment its trailing edge arrives, and compare two right triangles that share a side.

Step 1. Let the incident plane wavefront meet the mirror line at point $A$ first, while its other end $B$ is still a distance above the surface, poised to reach the mirror at a point $C$ farther along. The angle the incident wavefront makes with the mirror equals the angle of incidence $\theta_i$ (because the ray is perpendicular to the wavefront and the normal is perpendicular to the mirror, the two angles are equal).

Step 2. In the time $t$ it takes the trailing edge to travel from $B$ down to the mirror at $C$ , that leg has length $BC = vt$ . During the same time $t$ , the wavelet launched from $A$ spreads to a radius $AE = vt$ . Since both distances are covered at the same speed in the same time, $AE = BC = vt$ .

Step 3. The reflected wavefront is the tangent from $C$ to the wavelet centred on $A$ ; call the tangent point $E$ , so $CE$ is the reflected wavefront and $AE=vt$ is a radius meeting it at a right angle. Now compare triangles $ABC$ and $AEC$ : both are right-angled (at $B$ and at $E$ ), they share the hypotenuse $AC$ , and they have equal legs $BC = AE = vt$ . By the right-angle–hypotenuse–side congruence, the two triangles are congruent.

Step 4. Congruent triangles have equal corresponding angles, so $\angle BAC = \angle ECA$ . But $\angle BAC$ is the angle of incidence (wavefront-to-mirror on the way in) and $\angle ECA$ is the angle of reflection (wavefront-to-mirror on the way out). Therefore $\theta_i = \theta_r$ : a ray arriving at $30^\circ$ to the normal leaves at $30^\circ$ . The reflection law is a theorem of geometry once you accept the wavelet picture. Verified: with $\theta_i = 30^\circ$ the reflected angle is $\theta_r = 30^\circ$ (equal legs force congruent triangles) — recomputed in verify_hs2-huygens-principle.py.

Go deeper · Derivation B · 1 minSnell's law from the wavefront crossing a boundary

Derivation BSnell's law from the wavefront crossing a boundary

The goal is to obtain $\sin\theta_1/\sin\theta_2 = v_1/v_2 = n_2/n_1$ purely from Huygens' construction at a flat interface, showing that refraction is the geometric consequence of the wave travelling at different speeds on the two sides. Again we follow a single plane wavefront and build two right triangles that share the interface segment, then take a ratio of sines so the shared length cancels.

Step 1. Let the incident wavefront in medium 1 meet the interface at $A$ first, with its far edge $B$ reaching the surface at a point $C$ after a time $t$ . The distance along the incident wave is $BC = v_1 t$ , and the angle of incidence $\theta_1$ is the angle between the incident wavefront and the interface.

Step 2. During that same time $t$ , the wavelet that $A$ launches into medium 2 spreads at the slower speed $v_2$ , reaching only radius $AD = v_2 t$ . The refracted wavefront is the tangent $CD$ from $C$ to this smaller wavelet, and $\theta_2$ (the angle of refraction) is the angle between the refracted wavefront and the interface.

Step 3. Read the two right triangles that share the hypotenuse $AC$ (the stretch of interface between where the wave first lands and where its far edge lands). In triangle $ABC$ , $\sin\theta_1 = BC/AC = v_1 t / AC$ . In triangle $ACD$ , $\sin\theta_2 = AD/AC = v_2 t / AC$ .

Step 4. Divide the two: the shared $AC$ and $t$ cancel, leaving $\dfrac{\sin\theta_1}{\sin\theta_2} = \dfrac{v_1}{v_2}$ . Because $v=c/n$ , this is $\dfrac{v_1}{v_2} = \dfrac{c/n_1}{c/n_2} = \dfrac{n_2}{n_1}$ , i.e. $n_1\sin\theta_1 = n_2\sin\theta_2$ . For air into glass, $\sin\theta_2 = (n_1/n_2)\sin\theta_1 = 0.5/1.5 = 0.3333$ , so $\theta_2 = 19.47^\circ$ : the ray bends toward the normal on entering the slower, denser medium. The frequency rides through unchanged, so the wavelength shrinks to $\lambda_0/n = 400\ \mathrm{nm}$ . Verified: $\sin\theta_2 = 0.5/1.5 = 0.3333$ , $\theta_2 = 19.47^\circ$ , $v_{\text{glass}} = 2.0\times10^8\ \mathrm{m\,s^{-1}}$ , $\lambda_{\text{glass}} = 400\ \mathrm{nm}$ — recomputed in verify_hs2-huygens-principle.py.

Limiting cases

case	result	meaning
flat mirror, any $\theta_i$	$\theta_r = \theta_i$	equal wavelet radii force a symmetric bounce
air into glass, $n_2 > n_1$	ray bends toward the normal	wavelets slow on the denser side, the front tilts inward
glass into air, $n_2 < n_1$	ray bends away from the normal	the faster side lets the front tilt outward
normal incidence $\theta_1 = 0$	$\theta_2 = 0$ , no bend	equal-radius wavelets, but $\lambda$ still shrinks to $\lambda_0/n$
refraction, any medium	$f$ unchanged, $v=c/n$ , $\lambda=\lambda_0/n$	the source fixes frequency; speed and wavelength track $n$
opening $\gg \lambda$	straight-line (rectilinear) propagation	sideways wavelets cancel; only the forward envelope survives
opening $\sim \lambda$	the wave spreads (diffraction)	too few sources to cancel sideways

Common trapA wavefront is not a ray, and refraction does not change the frequency

Two confusions cost the most marks here. First, a wavefront is a surface of constant phase and a ray is the line perpendicular to it — they are not interchangeable, and Huygens' construction is built on wavefronts, not rays. Second, and more dangerous: when light enters glass or water, its frequency does not change. The source drives the boundary atoms at its own rate, so $f$ passes through untouched; what changes is the speed ( $v=c/n$ ) and, with it, the wavelength ( $\lambda=\lambda_0/n$ ). Writing "the frequency halves in glass" or "the colour changes underwater" is simply wrong — a red laser is still red under water. And remember that Huygens' principle is a construction, a recipe for finding the next wavefront, not a formula you plug numbers into; the numbers come from the reflection and Snell relations it produces.

Worked examples

Board level — refraction of a light beam into glass (CBSE and ICSE pattern)

A ray of monochromatic light of vacuum wavelength $600\ \mathrm{nm}$ strikes an air–glass surface at $30^\circ$ to the normal. The glass has refractive index $1.5$ . Find (a) the angle of refraction, (b) the speed of the light in the glass, (c) its wavelength in the glass, and (d) its frequency in the glass. Take $c = 3.0\times10^{8}\ \mathrm{m\,s^{-1}}$ .

Step 1. Angle of refraction from Snell's law: $\sin\theta_2 = (n_1/n_2)\sin\theta_1 = (1/1.5)\sin 30^\circ = 0.5/1.5 = 0.3333$ , so $\theta_2 = \boxed{19.47^\circ}$ . The ray bends toward the normal, as it must on entering a denser medium.

Step 2. Speed in glass: $v = c/n = (3.0\times10^{8})/1.5 = \boxed{2.0\times10^{8}\ \mathrm{m\,s^{-1}}}$ — two-thirds of the vacuum speed.

Step 3. Wavelength in glass: $\lambda_{\text{glass}} = \lambda_0/n = 600/1.5 = \boxed{400\ \mathrm{nm}}$ . The wave is physically shorter inside the glass.

Step 4. Frequency: it does not change on refraction, so compute it once from the vacuum values, $f = c/\lambda_0 = (3.0\times10^{8})/(600\times10^{-9}) = \boxed{5.0\times10^{14}\ \mathrm{Hz}}$ . Cross-check: $v/\lambda_{\text{glass}} = (2.0\times10^{8})/(400\times10^{-9}) = 5.0\times10^{14}\ \mathrm{Hz}$ — the same frequency the glass inherited from the source.

Competitive level — trusting the Huygens ratio (JEE and NEET pattern)

Using Huygens' construction, a student writes $\dfrac{\sin\theta_1}{\sin\theta_2} = \dfrac{v_1}{v_2}$ . For the air-to-glass case above, verify that this ratio equals $n_2/n_1$ , and state which way the wavelength and speed change. Then, without any new calculation, decide whether the frequency in the glass is greater than, equal to, or less than the frequency in air.

Step 1. Compute the speed ratio: $v_1/v_2 = c/(c/n_2) = n_2 = 1.5$ for air ( $n_1 = 1$ ) into glass. This equals $n_2/n_1 = 1.5/1 = \boxed{1.5}$ , confirming Huygens' construction reproduces Snell's law exactly.

Step 2. Direction of change: the denser medium is slower, so $v$ drops ( $3.0\times10^{8} \to 2.0\times10^{8}\ \mathrm{m\,s^{-1}}$ ) and $\lambda$ shrinks ( $600 \to 400\ \mathrm{nm}$ ), both by the factor $1/n = 2/3$ .

Step 3. Frequency comparison — the conceptual trap. Since the source sets $f$ and the boundary cannot alter it, the frequency in glass is $\boxed{\text{equal}}$ to the frequency in air, $5.0\times10^{14}\ \mathrm{Hz}$ . A tempting wrong answer is "it decreases because the light slowed down," but $f = v/\lambda$ and both $v$ and $\lambda$ fell by the same factor $2/3$ , so their ratio is unchanged. Timing note: about 60 seconds. The whole item is one Snell line plus the reflex that $f$ never changes on refraction — do not recompute $f$ separately for each medium; it is the one invariant.

Exam traps

Wavefront versus ray. A wavefront is a surface of constant phase; a ray is its perpendicular. Draw the wavelets on the wavefront, not on the ray, or the construction collapses. Questions that say "sketch the refracted wavefront" want the tilted tangent, not just a bent arrow.

Frequency is the invariant, not the wavelength. On refraction $f$ stays put while $v$ and $\lambda$ change together. The single most common slip is "the wavelength stays the same" or "the frequency changes" — it is the exact opposite. Anchor yourself to $f = v/\lambda$ : fix $f$ , and $v$ and $\lambda$ move in lock-step.

Which way does it bend? Into a denser medium (bigger $n$ , smaller $v$ ) the ray bends toward the normal; into a rarer medium it bends away. If your $\theta_2$ came out larger than $\theta_1$ for air-into-glass, you inverted the ratio — use $\sin\theta_2 = (n_1/n_2)\sin\theta_1$ .

Huygens is a construction, not a plug-in formula. There is no "Huygens equation" to substitute into. The principle produces the reflection and Snell relations; those are what you compute with. Marks are for the wavelet-and-envelope reasoning when a proof is asked for.

The suppressed backward wavelet. Huygens' bare principle would also send a wave backward, which is not observed. This is a genuine gap in the 1678 version; Fresnel and later Kirchhoff repaired it by adding interference and an obliquity factor. State that the backward wave is suppressed — do not try to derive why at this level.

Board scope. Huygens' construction and the wavefront derivations of reflection and Snell's law are assessed on CBSE, ICSE, JEE (Main and Advanced), and NEET. AP Physics 2, IB, and A-Level (CIE 9702) do not ask you to derive reflection or refraction from secondary wavelets — do not spend AP or IB time memorising this proof as an exam objective.

Origins · 1 minChristiaan Huygens and how this was discovered

Christiaan Huygens

1629–1695 · The Hague

Christiaan Huygens was a Dutch polymath — astronomer, mathematician, and the finest instrument-maker of his age — who built the first working pendulum clock, discovered Saturn's largest moon Titan, and correctly explained the true nature of its rings. For optics his monument is the Traité de la Lumière, written in 1678 and published in 1690, where he set out the wave picture of light and the wavelet construction this article is built on. From "every point of a wavefront is a new source, and the wave is their envelope," he derived reflection, refraction, and even the double refraction of Iceland spar. His wave theory lost ground for a century to Newton's rival corpuscular view, and was vindicated only when Thomas Young and Augustin-Jean Fresnel added interference — the missing ingredient — to Huygens' wavelets in the early 1800s.

History

1637

— René Descartes publishes the sine law of refraction in La Dioptrique, without a wave picture behind it.

1662

— Pierre de Fermat derives the same refraction law from his principle of least time.

1678

— Christiaan Huygens formulates the wavelet construction and reads it to the Académie des sciences in Paris.

1690

— Huygens publishes Traité de la Lumière, deriving reflection, refraction, and double refraction from secondary wavelets.

1704

— Isaac Newton's Opticks advances the rival corpuscular theory, and the wave view falls out of favour for a century.

1818

— Augustin-Jean Fresnel grafts interference onto Huygens' wavelets, curing the backward-wave flaw and founding the Huygens–Fresnel principle.

Going further

Texts. NCERT Physics Class 12, Chapter 10 ("Wave Optics"), the sections on Huygens' principle and the wavefront derivations of reflection and refraction · H. C. Verma, Concepts of Physics Vol. 1, "Light Waves" · Halliday, Resnick and Walker, the wave-optics chapter's treatment of wavefronts.

Provenance. Standard results. Recompute: $\sin\theta_2 = \sin 30^\circ/1.5 = 0.5/1.5 = 0.3333$ , $\theta_2 = 19.47^\circ$ ; $v_{\text{glass}} = 3.0\times10^{8}/1.5 = 2.0\times10^{8}\ \mathrm{m\,s^{-1}}$ (two-thirds of $c$ ); $\lambda_{\text{glass}} = 600/1.5 = 400\ \mathrm{nm}$ ; $f = 3.0\times10^{8}/600\times10^{-9} = 5.0\times10^{14}\ \mathrm{Hz}$ (unchanged); Huygens ratio $v_1/v_2 = n_2/n_1 = 1.5$ ; reflection $\theta_r = \theta_i = 30^\circ$ .

Prerequisites. Superposition & Interference of Waves — the addition rule that lets wavelets combine into an envelope. Leads to. Coherence & Interference: Path Difference, Phase & Intensity — where two Huygens sources are made to interfere · and, for enrichment, the university seam Huygens–Fresnel Principle, which repairs the backward-wave gap and turns the construction into the full theory of diffraction (go-deeper only, never an exam dependency here).

What exactly is the difference between a wavefront and a ray?

A wavefront is a surface — a line in a two-dimensional sketch — that joins all the points of a wave that are at the same phase at one instant, such as all the crests of a ripple. A ray is a single line drawn perpendicular to the wavefront, pointing in the direction the wave travels. They describe the same wave from two angles: the wavefront tells you where the wave is, the ray tells you which way it goes. Geometric optics prefers rays because they are easy to draw and bounce; wave optics prefers wavefronts because interference and diffraction are about phase, and phase is exactly what a wavefront tracks. Huygens' whole method lives on the wavefront side.

Why does the frequency stay the same but the wavelength change on refraction?

The frequency is set by the source and passed along like a handshake: the wave arriving at the glass surface wobbles the surface atoms at its own rate, and they in turn launch the wave inside the glass at that identical rate. Nothing at the boundary can speed up or slow down how many crests arrive per second, so $f$ is fixed. But inside the glass the wave travels slower, $v = c/n$ . Since $v = f\lambda$ and $f$ cannot change, the wavelength must fall to keep the product equal to the smaller speed: $\lambda = \lambda_0/n$ . Picture crests arriving at the same tempo but crawling forward more slowly — they bunch up, and "bunched-up crests" is a shorter wavelength.

If the light slows down in glass, why doesn't it get dimmer or lower in pitch?

"Pitch" is frequency, and that is exactly what does not change, so the colour is unchanged — a red laser stays red under water. Brightness is about the amplitude (how much energy per second the wave carries), not about the speed, so slowing down does not by itself dim the light. Some energy is reflected at the surface and a little is absorbed on the way through, which can dim a beam, but that is a separate effect from the refraction itself. The pure act of entering a denser medium changes only the speed and the wavelength, in the fixed ratio $1/n$ , while leaving frequency, and therefore colour, alone.

Why does light travel in straight lines if it is really a spreading wave?

Because everyday openings are enormous compared with light's wavelength of a few hundred nanometres. Huygens' wavelets spread from every point of the wavefront, but when the opening is far wider than $\lambda$ , the sideways-spreading wavelets from all those points interfere destructively and cancel, leaving only the straight-ahead envelope — so the wave marches in a straight line. Shrink the opening toward the wavelength and there are too few wavelet sources for the sideways cancellation to be complete; the wave then noticeably spreads, which is diffraction. So "light goes straight" is not a separate law but a large-aperture limit of the same wave behaviour that makes it bend through a narrow slit.

What is the backward-wave problem with Huygens' principle?

If every point of a wavefront is a full spherical source, each wavelet spreads backward as well as forward, so the bare 1678 principle predicts a wave travelling back toward the source — which is never seen. Huygens waved this away, and it remained a genuine flaw for well over a century. Fresnel fixed it in 1818 by insisting the wavelets interfere (not merely envelope), and by weighting them with an obliquity factor that is largest straight ahead and falls to zero backward; Kirchhoff later derived that factor from the wave equation. For this course you should know the backward wave is suppressed and that Fresnel supplied the cure, but you are not expected to derive the obliquity factor.

Does Huygens' principle work for sound and water waves too?

Yes — it is a statement about waves in general, not about light specifically. Any wave that obeys superposition can be advanced by treating each point of its wavefront as a secondary source and taking the envelope, so the construction describes sound bending around a doorway, water waves fanning out past a harbour gap, and seismic waves refracting between rock layers, just as well as light. The reflection and refraction laws it yields are the same in form; only the speed $v$ and what sets it differ from one kind of wave to another. This universality is why the method reappears across acoustics, seismology, and antenna engineering.

Is Snell's law really "derived" here, or just quoted?

It is genuinely derived, from the wavelet construction alone. The two right triangles built on the interface give $\sin\theta_1 = v_1 t / AC$ and $\sin\theta_2 = v_2 t / AC$ ; dividing them cancels the shared length and the time, leaving $\sin\theta_1/\sin\theta_2 = v_1/v_2$ , and $v = c/n$ turns that into $n_1\sin\theta_1 = n_2\sin\theta_2$ . No forces and no Maxwell's equations are needed — only "each point re-radiates at the local speed, and the new wavefront is the tangent." That is the beauty of Huygens' method and the reason it is still taught: a law that looks like it should need electromagnetism falls out of pure geometry plus one speed rule.