Ray Optics and Optical Instruments

The Cartesian Sign Convention 🔉✕⇢

Before we can derive a single mirror formula or a single lens formula that survives every geometric case, we must first agree on how distances are measured along the principal axis. The Cartesian sign convention supplies exactly this bookkeeping discipline. According to it, all distances are measured from one reference point: the pole of a spherical mirror or the optical centre of a lens. Nothing is measured from the centre of curvature, from the focus, or from the object itself. Fixing a single origin is the first act of self-consistency, because it guarantees that the object distance, the image distance, and the focal length are all reported against the same ruler laid along the principal axis. 🔉✕⇢

The second ingredient is a chosen positive direction. We take the direction in which the incident light actually travels, conventionally left to right, as the positive $x$ direction. Any distance measured in the same direction as the incident light is therefore counted positive, while any distance measured against the direction of the incident light is counted negative. This is why, for a real object sitting in front of a mirror, the light must travel from the object to the reflecting surface, and the object distance comes out negative: to reach the object from the pole you walk opposite to the incident beam. The sign is not an arbitrary decoration; it encodes geometry. 🔉✕⇢

Heights are handled by a companion rule. Perpendicular to the principal axis, distances measured upward from the axis are taken as positive and distances measured downward are taken as negative. Thus an erect object of height $h$ is a positive number, and an inverted image of the same physical size carries a negative height $h'$. Because the magnification is defined as $m=h'/h$, this height convention is precisely what lets a single number report both the size ratio and the orientation of the image in one stroke, a point we will exploit when the mirror and lens magnifications are derived. 🔉✕⇢

It is worth stressing why one insists on a convention at all. Without it, every geometric configuration, an object beyond the centre of curvature, an object between the focus and the pole, a concave mirror, a convex mirror, would demand its own separately memorised formula with its own placement of plus and minus signs. That is a recipe for error under examination pressure. With one accepted convention, it turns out that a single formula for spherical mirrors and a single formula for spherical lenses can handle all the different cases. The convention buys generality: you learn one relation and let the signs do the case analysis for you. 🔉✕⇢

Consider how the convention assigns signs to the standard optical elements. For a concave mirror the focus and the centre of curvature lie in front of the mirror, on the same side as the incident light source, so both the focal length and the radius of curvature are negative. For a convex mirror the focus and the centre of curvature lie behind the reflecting surface, away from the incident light, so both the focal length and the radius of curvature are positive. These signs are consequences of the convention, not extra facts to remember, and the relation $f=R/2$ carries the sign automatically. 🔉✕⇢

The same logic disciplines the image distance. If the reflected rays actually converge and meet in front of the mirror, the image is real and lies on the incident-light side, so its distance is negative. If the reflected rays only appear to diverge from a point behind the mirror, the image is virtual and its distance is positive. Thus the mere sign of the computed image distance already tells you the nature of the image before you draw a single ray. A negative image distance for a mirror announces a real, and for a mirror also inverted, image; a positive one announces a virtual, erect image. 🔉✕⇢

A frequent source of confusion is the belief that the convention changes depending on whether the mirror is concave or convex, or whether the object is real or virtual. It does not. The convention is a fixed coordinate frame anchored at the pole with the incident light defining the positive axis. What changes from case to case are the numerical values and their resulting signs, not the rules that assign them. This invariance is exactly what makes the convention powerful: you set up the same axis every time, substitute the signed quantities, and read off a signed answer whose sign is itself physically meaningful. 🔉✕⇢

The convention also clarifies the treatment of a virtual object, a subtlety that appears in multi-element systems such as two lenses in contact. When converging rays are intercepted by a second surface before they meet, the point toward which they were heading acts as an object lying on the far side, measured in the direction of the incident light, and therefore carries a positive object distance. Beginners who apply the convention mechanically, measuring every distance from the same pole with the same positive direction, handle this case correctly, whereas those who memorise sign patterns for real objects alone stumble. 🔉✕⇢

It helps to rehearse the bookkeeping on a concrete concave mirror. Suppose an object stands $20\ \text{cm}$ in front of a concave mirror whose radius of curvature is $30\ \text{cm}$. By the convention the object distance is $u=-20\ \text{cm}$, the radius is $R=-30\ \text{cm}$, and hence the focal length is $f=R/2=-15\ \text{cm}$. Every quantity that lies on the incident-light side of the pole has picked up a minus sign purely from the geometry. When these signed numbers are fed into the mirror equation, the computed image distance emerges with a sign that we simply interpret, rather than argue about after the fact. 🔉✕⇢

Finally, note the deep economy the convention delivers across the whole chapter. The very same sign rules that govern reflection by spherical mirrors also govern refraction at a spherical surface and refraction by thin lenses; only the physical formula changes, never the coordinate discipline. Because distances measured with the incident light are positive and those against it are negative, and because heights above the axis are positive, one consistent frame threads reflection and refraction together. Mastering this convention is therefore not a preliminary chore but the single most reusable skill in ray optics, since $u$, $v$, $f$, and $R$ never again need case-by-case sign guessing. 🔉✕⇢

In practice a reliable examination habit is to draw the principal axis, mark the pole as the origin, draw an arrow showing the incident light as the positive direction, and only then translate each physical distance into a signed number. If a computed value such as an image distance satisfies $v\lt 0$ for a mirror, you immediately know the image is real and on the object side; if $v\gt 0$, the image is virtual and behind the mirror. Trusting the signs, rather than overriding them with intuition, is what separates a clean solution from a sign-error-riddled one under time pressure. 🔉✕⇢

Reflection by Spherical Mirrors & Focal Length 🔉✕⇢

Reflection by spherical mirrors rests on the same two laws of reflection that govern any reflecting surface. The angle of reflection, measured between the reflected ray and the normal, equals the angle of incidence, measured between the incident ray and the same normal; and the incident ray, the reflected ray, and the normal at the point of incidence all lie in one plane. What makes the spherical case special is the identity of the normal. For a curved reflecting surface the normal is taken along the radius, that is, along the line joining the centre of curvature of the mirror to the point of incidence, because the normal must be perpendicular to the tangent to the surface at that point. 🔉✕⇢

A spherical mirror carries a small vocabulary that must be used precisely. The geometric centre of the mirror is called its pole, usually labelled $P$. The centre of the sphere of which the reflecting surface is a part is the centre of curvature $C$, and the distance $PC$ is the radius of curvature $R$. The straight line joining the pole and the centre of curvature is the principal axis, the axis of symmetry about which the whole optical behaviour is organised. Every distance we later measure, object distance, image distance, focal length, is measured from the pole along this principal axis, which is exactly what the sign convention demands. 🔉✕⇢

Now send a parallel beam of light, travelling close to and nearly along the principal axis, onto the mirror. For a concave mirror the reflected rays actually converge and cross at a single point $F$ on the principal axis. For a convex mirror the reflected rays diverge outward but, when produced backward, appear to come from a single point $F$ behind the mirror. In both cases this point is the principal focus of the mirror, and the distance from the pole $P$ to the focus $F$ is the focal length $f$. The concave mirror thus has a real focus in front of it, while the convex mirror has a virtual focus behind it. 🔉✕⇢

The crucial quantitative fact is that for such a beam the focal length is exactly half the radius of curvature, $f=R/2$. This is not an empirical accident; it follows from the laws of reflection applied to a ray parallel to the principal axis, together with the small-angle behaviour of the paraxial regime. Because a single number $R$ then fixes $f$, and because the sign convention makes both negative for a concave mirror and both positive for a convex mirror, the relation $f=R/2$ is one of the most heavily used shortcuts in the entire chapter and deserves to be understood by derivation rather than merely memorised. 🔉✕⇢

The word paraxial carries real weight and must not be skipped over. A paraxial ray is one incident at a point close to the pole and making a small angle with the principal axis. Only for such rays do the reflected rays from a parallel beam meet at a single sharp focus, because only then are the small-angle approximations, in which the tangent of an angle is replaced by the angle itself, accurate. The result $f=R/2$ is therefore a paraxial result. It is the clean idealisation on which the mirror equation is subsequently built, and its limitations are physical, not mathematical carelessness. 🔉✕⇢

The aperture of a mirror, its lateral size or width, is what decides whether the paraxial assumption is safe. If the aperture is small compared with the radius of curvature, every ray that strikes the mirror is nearly paraxial and the beam focuses tightly. As the aperture grows, rays striking the outer zones of the mirror are far from the pole and make appreciable angles with the axis, so the small-angle approximation breaks down for them. This is why textbook derivations quietly assume a small aperture: it is the condition under which a spherical mirror behaves like an ideal focusing element with a well-defined single focus. 🔉✕⇢

When the aperture is not small, the failure of the paraxial approximation shows up as spherical aberration. Rays reflected from the outer zones of a wide concave mirror cross the axis closer to the mirror than the paraxial rays do, so instead of a single sharp focus one obtains a blurred region and a bright envelope known as the caustic. Spherical aberration is a defect of the spherical shape itself, not of the material or polish, which is precisely why large telescope objectives favour a paraboloidal mirror: a parabola brings a genuinely parallel axial beam to one exact focus, with no aberration of this kind. 🔉✕⇢

The contrasting behaviour of concave and convex mirrors follows directly from where the focus sits. A concave mirror, curving inward toward the incoming light, is a converging mirror: parallel rays are brought together at a real focus, and depending on the object position it can form real or virtual images. A convex mirror, curving away from the incoming light, is a diverging mirror: parallel rays spread apart and only appear to come from a virtual focus behind the mirror. This is why a convex mirror always yields an erect, diminished, virtual image and is chosen for wide-view rear and side mirrors on vehicles. 🔉✕⇢

It is illuminating to note how many rays actually take part. An infinite number of rays leave every point of an object in all directions, and the laws of reflection hold at each and every point of the reflecting surface. A point is a genuine image point only if every ray leaving the corresponding object point passes through it after reflection. This is also why covering part of a mirror does not chop the image in half: the uncovered portion still receives rays from the whole object and forms a complete, though dimmer, image, since the intensity, not the extent, of the image depends on the reflecting area. 🔉✕⇢

Two convenient rays make ray tracing for spherical mirrors quick. A ray from an object point travelling parallel to the principal axis reflects so as to pass through the focus of a concave mirror, or to appear to come from the focus of a convex mirror. A ray directed through, or toward, the centre of curvature strikes the mirror along its own normal and simply retraces its path. The intersection of any two such reflected rays locates the image point. These constructions, combined with the laws of reflection and the relation $f=R/2$, are the geometric backbone from which the algebraic mirror equation is derived. 🔉✕⇢

Pulling the strands together, reflection by a spherical mirror is completely specified once you know the laws of reflection, the geometry of pole, centre of curvature, principal axis, and principal focus, and the paraxial relation $f=R/2$. The concave mirror converges light to a real focus while the convex mirror diverges it from a virtual focus, and the small-aperture paraxial condition is what keeps that focus sharp. Everything that follows, the mirror equation and the magnification, is a quantitative reformulation of this geometric picture under the Cartesian sign convention, so a firm grasp here pays off across every image-formation problem. 🔉✕⇢

The Mirror Equation & Magnification 🔉✕⇢

The mirror equation is the algebraic statement of image formation by a spherical mirror. It relates three signed quantities, the object distance $u$, the image distance $v$, and the focal length $f$, all measured from the pole along the principal axis, through the compact relation $\frac{1}{v}+\frac{1}{u}=\frac{1}{f}$. What makes this single equation so valuable is that, once the Cartesian sign convention is respected, it applies to concave and convex mirrors alike and to real and virtual images alike. There is no separate formula to memorise for each configuration; the signs of the numbers carry all the case-by-case information, which is exactly the economy the convention was designed to deliver. 🔉✕⇢

The derivation begins from a ray diagram for a concave mirror forming a real, inverted image, and it rests entirely on similar triangles. When an object $AB$ stands on the principal axis and its image $A'B'$ is formed after reflection, two pairs of triangles turn out to be similar: one pair built from the ray through the focus and one pair built from the ray reflecting at the pole. Ratios of corresponding sides of these similar triangles give relations among the magnitudes of the distances, and the sign convention is then imposed at the very end to convert those magnitude relations into the signed mirror equation. 🔉✕⇢

Linear magnification measures how large the image is relative to the object, and it is defined as the ratio of the height of the image to the height of the object, $m=\frac{h'}{h}$. Using the triangle formed by the ray incident at the pole, where the incident and reflected rays make equal angles with the principal axis, one finds that this height ratio equals $-\frac{v}{u}$. Hence the working formula $m=-\frac{v}{u}=\frac{h'}{h}$. Because heights above the axis are positive and those below are negative under the convention, the single number $m$ reports both how much the image is enlarged or diminished and whether it is erect or inverted. 🔉✕⇢

The sign of $m$ is therefore loaded with physical meaning. A negative magnification means the image height $h'$ has the opposite sign to the object height $h$, so the image is inverted; for a mirror this always accompanies a real image, whose reflected rays actually converge in front of the mirror. A positive magnification means the image is erect, which for a mirror accompanies a virtual image formed behind the mirror by rays that only appear to diverge. Thus the two numbers you compute, the signed image distance $v$ and the signed magnification $m$, together announce the position, the nature, and the orientation of the image without any further ray drawing. 🔉✕⇢

The magnitude of $m$ separates enlargement from diminution. If the magnitude of $m$ exceeds one the image is magnified, if it is less than one the image is diminished, and if it equals one the image is the same size as the object. Combining magnitude and sign lets a single computed value summarise the whole outcome: for instance $m=-3$ describes an image three times as tall as the object and inverted, hence real, while $m=+\tfrac{1}{2}$ describes an upright image half the object's height, hence virtual. Learning to read $m$ fluently is one of the quickest routes to correct answers under time pressure. 🔉✕⇢

Consider how the image of a concave mirror evolves as the object moves inward from far away. With the object beyond the centre of curvature, the image is real, inverted, and diminished, forming between the focus and the centre of curvature. As the object reaches the centre of curvature, the image is real, inverted, and the same size, formed at the centre of curvature itself. Bringing the object between the centre of curvature and the focus produces a real, inverted, magnified image lying beyond the centre of curvature. In each stage the image distance $v$ computed from the mirror equation comes out negative, signalling a real image on the object side. 🔉✕⇢

The behaviour changes character once the object crosses the focus of the concave mirror. With the object placed between the focus and the pole, the reflected rays diverge and no longer meet in front of the mirror; they only appear to come from a point behind it. The mirror equation then returns a positive image distance $v$, marking a virtual image, and the magnification turns positive and greater than one, so the image is erect and magnified. This is the shaving-mirror or make-up-mirror regime, in which a concave mirror close to the face yields an enlarged upright view, a direct algebraic consequence of the object lying inside the focal length. 🔉✕⇢

A convex mirror behaves far more simply, and the mirror equation makes the simplicity transparent. Its focal length is positive, and for any real object with negative $u$ the equation $\frac{1}{v}=\frac{1}{f}-\frac{1}{u}$ forces $v$ to be positive, so the image is always virtual and located behind the mirror. The magnification $m=-\frac{v}{u}$ then always comes out positive and less than one, so the image is invariably erect and diminished, regardless of where the object sits. This is precisely why convex mirrors are chosen as vehicle rear-view and side mirrors: they give an upright, reduced image over a wide field of view. 🔉✕⇢

It is worth appreciating that the very same equation, derived from a concave mirror forming a real image, silently handles all these outcomes because the algebra respects the signs. You never decide in advance whether an image will be real or virtual; you substitute the signed $u$ and $f$, solve for $v$, and then read the sign. A negative $v$ for a mirror is a real image on the incident-light side; a positive $v$ is a virtual image behind the mirror. The equation is, in this sense, a small case-analysis engine that you feed with signed inputs and query with signed outputs. 🔉✕⇢

There is also a subtle relationship between the mirror equation and the paraxial focal relation $f=R/2$. The mirror equation presumes a well-defined single focal length, which itself only exists in the paraxial, small-aperture regime where a parallel beam converges to one point. For wide apertures the focus smears into a caustic through spherical aberration, and the notion of a single $v$ satisfying the mirror equation becomes an approximation. In routine problem solving the paraxial assumption is taken for granted, but it is worth remembering that the clean predictive power of $\frac{1}{v}+\frac{1}{u}=\frac{1}{f}$ is inherited from that same idealisation. 🔉✕⇢

To use these tools reliably, adopt a fixed routine. Write down $f$ with its correct sign, negative for a concave mirror and positive for a convex mirror; write $u$ as a negative number for a real object; solve $\frac{1}{v}=\frac{1}{f}-\frac{1}{u}$ for $v$; then compute $m=-\frac{v}{u}$. Interpret the results by their signs: $v\lt 0$ means a real, inverted image and $v\gt 0$ means a virtual, erect one, while the magnitude of $m$ compared with one settles enlargement. Trusting this signed output, rather than overriding it with expectation, is what turns the mirror equation into a dependable instrument. 🔉✕⇢

Refraction & Snell's Law 🔉✕⇢

Refraction is the phenomenon that occurs when a beam of light encounters another transparent medium at an interface. At the boundary a part of the light gets reflected back into the first medium while the rest enters the other. A ray of light represents such a beam, and the direction of propagation of an obliquely incident ray, with an angle of incidence between $0^\circ$ and $90^\circ$, changes as it crosses the interface between the two media. This bending of the ray at the surface separating two transparent media is what we call refraction of light. It arises because light travels with different speeds in different media, and it is governed by two simple, experimentally established laws first quantified by Snell. Note that if the ray strikes the interface normally, along the normal itself, it passes straight through without any bending; only an obliquely incident ray is deviated, and it is this oblique refraction that produces the familiar sight of a straight object appearing broken at a water surface. Throughout this discussion we treat light in the ray picture, tracing the incident ray and the refracted ray on either side of the point of incidence. 🔉✕⇢

Full derivation, worked example and interactive 3D on the Refraction & Snell's Law tab →

Total Internal Reflection & Optical Fibres 🔉✕⇢

When a ray of light travels from an optically denser medium to a rarer medium, at the interface it is partly reflected back into the same denser medium and partly refracted into the rarer medium. This partial reflection into the originating medium is called internal reflection. Because the second medium is rarer, the refracted ray bends away from the normal, so the angle of refraction $r$ is larger than the angle of incidence $i$. As long as $i$ is modest, both a reflected ray and a refracted ray coexist, and the refracted ray carries away most of the light energy while the internally reflected ray remains comparatively feeble. This everyday situation, light escaping upward from water into air, is the starting point for understanding total internal reflection. 🔉✕⇢

Full derivation, worked example and interactive 3D on the Total Internal Reflection & Optical Fibres tab →

Refraction at a Spherical Surface 🔉✕⇢

So far we have treated refraction only at a plane interface, where the two transparent media meet along a flat boundary. We now consider refraction at a single spherical surface separating two media of refractive index $n_1$ and $n_2$. An infinitesimal patch of a spherical surface can be regarded as planar, so the same laws of refraction apply at every point on the surface. Just as for a spherical mirror, the normal at the point of incidence is perpendicular to the tangent plane at that point and therefore passes through the centre of curvature $C$. This single refracting surface is the fundamental building block from which the lens maker's formula is later assembled. 🔉✕⇢

Consider a point object $O$ on the principal axis, lying in the medium of refractive index $n_1$. Light from $O$ strikes a spherical surface of radius of curvature $R$ whose centre of curvature is $C$, and refracts into the second medium of refractive index $n_2$, forming the image $I$ on the axis. A ray from $O$ meets the surface at a point $N$ close to the pole $M$, bends according to Snell's law, and proceeds toward $I$. We take the aperture, or the lateral size of the surface, to be small compared with the object and image distances, so that every ray remains paraxial and makes only a small angle with the principal axis. 🔉✕⇢

The paraxial approximation is the engine of the whole derivation. Because each ray stays near the axis, the point $N$ lies close to the pole $M$, and the perpendicular distance $MN$ is nearly equal to the arc height. For any small angle $\theta$ we may replace $\tan\theta$ and $\sin\theta$ by $\theta$ itself, since the correction terms are of higher order and negligible. This linearisation converts the awkward trigonometry of a curved refracting surface into simple ratios of lengths, exactly as it did for the focal length of a spherical mirror. Without this restriction the image would suffer aberration and no single sharp image point would exist. 🔉✕⇢

We now write the three relevant small angles as ratios. For the ray from the object, $\tan(\angle NOM)=\dfrac{MN}{OM}$; for the line to the centre of curvature, $\tan(\angle NCM)=\dfrac{MN}{MC}$; and for the ray reaching the image, $\tan(\angle NIM)=\dfrac{MN}{MI}$. The key geometric insight is the exterior-angle theorem: in triangle $NOC$ the angle of incidence $i$ is the exterior angle at $N$, so it equals the sum of the two remote interior angles, $i=\angle NOM+\angle NCM$. This clean relationship is what makes the spherical-surface derivation so much simpler than a brute-force ray trace. 🔉✕⇢

Applying the exterior-angle result and the small-angle ratios, the angle of incidence becomes $i=\dfrac{MN}{OM}+\dfrac{MN}{MC}$. For the refracted ray a similar triangle gives the angle of refraction as a difference of angles, $r=\angle NCM-\angle NIM=\dfrac{MN}{MC}-\dfrac{MN}{MI}$. Both $i$ and $r$ are now expressed purely through the perpendicular height $MN$ and the three axial distances $OM$, $MC$ and $MI$. Notice that the common factor $MN$ will cancel presently, which is why the final relation between object distance and image distance is independent of exactly where on the small aperture the ray happened to strike. 🔉✕⇢

The two media are linked by Snell's law, $n_1\sin i=n_2\sin r$. In the paraxial regime the sines reduce to the angles themselves, so this becomes simply $n_1 i=n_2 r$. Substituting the expressions for $i$ and $r$ and cancelling the common height $MN$ yields a relation among the magnitudes $OM$, $MI$ and $MC$: $\dfrac{n_1}{OM}+\dfrac{n_2}{MI}=\dfrac{n_2-n_1}{MC}$. At this stage every length is still a positive magnitude, because we have not yet decided on directions. The physics of refraction is complete; what remains is only the bookkeeping of signs so that one formula can serve every geometric case. 🔉✕⇢

We now impose the Cartesian sign convention consistently. All distances are measured from the pole of the surface. Distances measured in the same direction as the incident light are taken as positive, and those measured against it are negative. Here light travels from the object toward the surface, so to reach the object we travel opposite to the incident light, giving $OM=-u$. The image lies in the direction of the incident light, so $MI=+v$, and for the convex-toward-the-object surface shown the centre of curvature also lies that way, so $MC=+R$. Heights above the principal axis are positive and those below are negative, matching the mirror convention. 🔉✕⇢

Substituting $OM=-u$, $MI=+v$ and $MC=+R$ into the magnitude relation gives the compact and famous result $\dfrac{n_2}{v}-\dfrac{n_1}{u}=\dfrac{n_2-n_1}{R}$. This equation relates the object distance and the image distance to the refractive index of each medium and to the radius of curvature of the single spherical surface. It is strikingly general: nowhere did we assume the surface bulged one particular way. The same formula therefore governs a surface that is convex toward the incoming light and one that is concave toward it; only the algebraic sign of $R$ changes, and the mathematics automatically produces the correct real or virtual image. 🔉✕⇢

The sign of $R$ is fixed by the location of the centre of curvature relative to the pole. If $C$ lies on the outgoing side, in the direction of the incident light, then $R\gt 0$; if $C$ lies on the incoming side, then $R\lt 0$. Likewise a positive $v$ denotes a real image formed on the far side, while a negative $v$ denotes a virtual image on the same side as the object. Because the derivation nowhere assumed a real image, you may verify that the identical relation holds when the refracted rays only appear to diverge, provided the sign convention is applied faithfully throughout. 🔉✕⇢

This one relation for a single spherical surface is the seed of nearly all lens theory. A thin lens is a transparent medium bounded by two surfaces, at least one of which is spherical. By applying $\dfrac{n_2}{v}-\dfrac{n_1}{u}=\dfrac{n_2-n_1}{R}$ at the first surface, treating its image as the object for the second, and then adding the two results in the thin-lens limit, one obtains the lens maker's formula and thereafter the thin lens formula. Mastering the paraxial approximation, the exterior-angle geometry and the Cartesian sign convention here therefore pays dividends across every optical instrument built from lenses. 🔉✕⇢

A useful sanity check is the limiting behaviour of the formula. If the two media have equal refractive index, $n_2=n_1$, the right-hand side vanishes and the surface bends nothing; the object and image distances satisfy $n/v=n/u$, i.e. no net refraction, exactly as a submerged lens of matching index would predict. Similarly, letting $R\to\infty$ turns the spherical surface into a plane interface, recovering the apparent-depth result for near-normal viewing. These consistency checks, together with the units, are quick tools in an examination to confirm that signs and reciprocals of the refractive index have been entered correctly before any arithmetic is attempted. 🔉✕⇢

The Lens Maker's Formula 🔉✕⇢

A thin lens is a transparent optical medium bounded by two surfaces, at least one of which is spherical. Having derived the relation for refraction at a single spherical surface, we can now build the lens maker's formula simply by applying that relation twice, once at each face, and adding the results in the thin-lens limit. The formula we shall obtain, $\dfrac{1}{f}=(n_{21}-1)\left(\dfrac{1}{R_1}-\dfrac{1}{R_2}\right)$, connects the focal length of a lens to the two radii of curvature and to the refractive index of the lens material relative to its surroundings. It is the practical recipe an optician or manufacturer uses to grind lenses of any desired focal length. 🔉✕⇢

Consider a double convex lens of material refractive index $n_2$ immersed in a medium of refractive index $n_1$. Light from an object $O$ first refracts at the front surface of radius $R_1$, which would by itself form an intermediate image $I_1$. This image then serves as the object for the second surface of radius $R_2$, which refracts the light once more to produce the final image $I$. The whole action of the lens is thus the composition of two single-surface refractions. Because the lens is thin, the two surfaces are treated as essentially coincident at the optical centre, and the small thickness between them is neglected throughout. 🔉✕⇢

Apply the single-surface relation to the first interface, where light passes from medium $n_1$ into the glass $n_2$. Writing the object distance as $OB$ and the intermediate image distance as $BI_1$, we obtain $\dfrac{n_2}{BI_1}-\dfrac{n_1}{OB}=\dfrac{n_2-n_1}{R_1}$. This is nothing more than the spherical-surface formula with the appropriate media on each side. The intermediate image $I_1$ may be real or virtual, but its position is dictated entirely by this first refraction; it is only a stepping stone to the final image and never needs to be physically formed on a screen. 🔉✕⇢

Now treat the second interface, where light emerges from the glass $n_2$ back into the surrounding medium $n_1$. Here the roles of the two refractive index values are exchanged, and the intermediate image $I_1$ acts as the object. Applying the single-surface relation again gives $\dfrac{n_1}{DI}-\dfrac{n_2}{DI_1}=\dfrac{n_1-n_2}{R_2}$. In the thin-lens approximation the points $B$ and $D$ both collapse onto the optical centre, so the intermediate image distances satisfy $BI_1=DI_1$. This coincidence is exactly what allows the two equations to be combined cleanly, with the intermediate term cancelling in the sum. 🔉✕⇢

Adding the two surface equations, the intermediate image distance drops out entirely, leaving $\dfrac{n_1}{DI}-\dfrac{n_1}{OB}=(n_2-n_1)\left(\dfrac{1}{R_1}+\dfrac{1}{R_2}\right)$ once the algebra of signs on $R_2$ is respected. Dividing through by $n_1$ and using the relative refractive index $n_{21}=n_2/n_1$ tidies the expression. The remarkable feature is that the intermediate image $I_1$, which was central to the physical picture, has vanished from the final result; only the object distance, the final image distance, the two radii of curvature and the relative index survive. 🔉✕⇢

To extract the focal length we send the object to infinity, so that $OB\to\infty$ and the final image forms at the focus, $DI=f$. The reciprocal object term then vanishes, and we are left with $\dfrac{1}{f}=(n_{21}-1)\left(\dfrac{1}{R_1}-\dfrac{1}{R_2}\right)$ after the sign convention converts the surface radii into their signed forms $R_1$ and $R_2$. This is the lens maker's formula. Combined with the accompanying result $\dfrac{1}{v}-\dfrac{1}{u}=\dfrac{1}{f}$, it lets us predict image formation for any thin lens once its geometry and material are known. 🔉✕⇢

The sign convention on the two radii is what distinguishes a converging lens from a diverging one. For the double convex lens the centre of curvature of the first surface lies on the outgoing side, giving $R_1\gt 0$, while the centre of curvature of the second surface lies on the incoming side, giving $R_2\lt 0$. With $n_{21}\gt 1$ the bracket $\left(\dfrac{1}{R_1}-\dfrac{1}{R_2}\right)$ is positive, so $f\gt 0$ and the lens converges light. A double concave lens reverses both radii, making $f\lt 0$, so it diverges the beam. The single formula thus handles convex and concave lenses alike. 🔉✕⇢

It is instructive to see how the shape maps onto the signs. A convex surface met by the incident light contributes a positive radius, a concave surface a negative one; the second face contributes with the opposite geometric sense because its centre of curvature sits on the other side. Feeding these signed radii into the bracket automatically yields a positive focal length for any net-converging shape and a negative focal length for any net-diverging shape. There is no need to memorise separate formulae for biconvex, plano-convex, or meniscus lenses; the Cartesian convention encodes every case in one compact expression. 🔉✕⇢

The lens maker's formula also reveals a strong dependence on the surrounding medium, through the factor $n_{21}-1$, where $n_{21}$ is the refractive index of the lens material relative to its surroundings. A glass lens that converges strongly in air converges far more weakly in water, because $n_{21}$ falls closer to unity and the bracket is multiplied by a smaller number, lengthening the focal length. In the striking limiting case where the lens material and the surrounding liquid share the same refractive index, $n_{21}=1$, the factor vanishes, $1/f=0$, and the lens becomes optically invisible, behaving like a flat sheet with no converging or diverging power at all. 🔉✕⇢

This medium dependence is not a mere curiosity; it underlies the classic demonstration of a glass lens vanishing when lowered into a liquid of matching index. It also explains why the focal length quoted for a lens is always tied to a stated surrounding medium, usually air. When solving problems you must recompute $n_{21}$ for the actual surroundings before using the formula: a lens of focal length $20\text{ cm}$ in air can acquire a focal length of nearly $78\text{ cm}$ in water. Overlooking this shift in focal length is one of the most common sources of error in lens combination and optical instrument problems, and examiners deliberately test it by immersing an otherwise familiar glass lens in a liquid of stated refractive index. 🔉✕⇢

Finally, note the assumptions baked into the derivation. The lens is thin, so its thickness is negligible and the two surfaces are coincident; the rays are paraxial, so aberrations are ignored; and the same relative refractive index describes both refractions. Real thick lenses give slightly coloured images through dispersion, since $n_{21}$ varies with wavelength, and multi-component designs are used in cameras, microscopes and telescopes to correct such defects. The thin lens maker's formula is nonetheless the indispensable first approximation, and every more sophisticated lens-design calculation is built upon this same spherical-surface foundation. 🔉✕⇢

The Thin Lens Formula & Magnification 🔉✕⇢

A thin lens is a transparent optical medium bounded by two refracting surfaces, at least one of which is spherical, whose thickness is negligible compared with the object distance, the image distance and the radii of curvature involved. Because the thickness can be ignored, both refracting surfaces are treated as passing through a single point on the principal axis called the optical centre. The line through the optical centre and the two centres of curvature is the principal axis. When paraxial rays (rays close to the axis making small angles with it) from a point object are refracted by such a lens, they reconverge to, or appear to diverge from, a single image point. The single relation that ties the object distance $u$, the image distance $v$ and the focal length $f$ together is the thin lens formula $\frac{1}{v}-\frac{1}{u}=\frac{1}{f}$, and it holds for every situation once the sign convention is respected. 🔉✕⇢

Full derivation, worked example and interactive 3D on the The Thin Lens Formula & Magnification tab →

Power of a Lens & Combination of Lenses 🔉✕⇢

The power of a lens is a measure of the convergence or divergence that the lens introduces into a beam of light falling on it. A lens of shorter focal length bends the incident rays more sharply, converging them in the case of a convex lens and diverging them in the case of a concave lens. Quantitatively the power $P$ is defined as the tangent of the angle by which the lens deflects a ray that arrives parallel to the principal axis at unit distance from the optical centre. For small deflections this reduces to the beautifully simple relation $P=\dfrac{1}{f}$, tying the power directly to the reciprocal of the focal length. 🔉✕⇢

The SI unit of power is the dioptre, written $\text{D}$, defined by $1\,\text{D}=1\,\text{m}^{-1}$. A lens whose focal length is one metre therefore has a power of one dioptre. Because power is the reciprocal of focal length, a short-focal-length lens is a high-power lens. The sign of the power follows the sign of the focal length under the Cartesian convention: a converging convex lens has $f\gt 0$ and hence $P\gt 0$, while a diverging concave lens has $f\lt 0$ and hence $P\lt 0$. This signed quantity is exactly what an optician manipulates when prescribing corrective lenses. 🔉✕⇢

A concrete reading of the dioptre makes the idea tangible. When an optician prescribes a corrective lens of power $+2.5\,\text{D}$, the required lens is a convex lens of focal length $f=\dfrac{1}{2.5}\,\text{m}=+40\,\text{cm}$. Conversely a prescription of $-4.0\,\text{D}$ denotes a concave lens of focal length $-25\,\text{cm}$. The larger the magnitude of the power, the more strongly the lens bends light and the shorter its focal length. Expressing lens strength in dioptres is convenient precisely because, as we shall see, the powers of lenses placed in contact simply add as signed numbers. 🔉✕⇢

Consider now two thin lenses $A$ and $B$ of focal lengths $f_1$ and $f_2$ placed in contact with each other, their optical centres taken to be coincident at a common point $P$ since the lenses are thin. Let an object be placed beyond the focus of the first lens. Lens $A$ alone would form an image $I_1$, which then acts as the object for lens $B$, producing the final image $I$. The intermediate image $I_1$ need not be physically realised; it is only a device for locating the final image. This two-step picture is the same compositional idea used to derive the lens maker's formula. 🔉✕⇢

For the first lens the thin lens formula gives $\dfrac{1}{v_1}-\dfrac{1}{u}=\dfrac{1}{f_1}$, and for the second lens, taking $I_1$ as its object, $\dfrac{1}{v}-\dfrac{1}{v_1}=\dfrac{1}{f_2}$. Adding these two equations, the intermediate image distance $v_1$ cancels neatly, leaving $\dfrac{1}{v}-\dfrac{1}{u}=\dfrac{1}{f_1}+\dfrac{1}{f_2}$. If the combination is regarded as a single equivalent lens of focal length $f$, then $\dfrac{1}{v}-\dfrac{1}{u}=\dfrac{1}{f}$, and comparing the two expressions yields the central result $\dfrac{1}{f}=\dfrac{1}{f_1}+\dfrac{1}{f_2}$ for two thin lenses in contact. 🔉✕⇢

The derivation extends immediately to any number of thin lenses in contact. For lenses of focal length $f_1,f_2,f_3,\dots$ stacked together, the effective focal length of the combination is given by $\dfrac{1}{f}=\dfrac{1}{f_1}+\dfrac{1}{f_2}+\dfrac{1}{f_3}+\cdots$. Because power is the reciprocal of focal length, this reciprocal sum becomes an ordinary sum of powers: $P=P_1+P_2+P_3+\cdots$. The powers add as a simple algebraic sum, in which some terms may be positive for convex lenses and others negative for concave lenses. This additivity is the single most useful property of the dioptre in practical optics. 🔉✕⇢

The additive rule makes lens design intuitive. Placing a convex lens of power $+5\,\text{D}$ in contact with a concave lens of power $-3\,\text{D}$ gives a net power of $+2\,\text{D}$, a weaker converging combination, without any lengthy focal-length arithmetic. Because the sum is algebraic, a strong convex lens can be trimmed by a weaker concave lens, or a diverging system built from an excess of concave power. Designers exploit this to reach a target power or focal length that no single available lens provides, simply by combining stock lenses of known dioptric strength in contact. 🔉✕⇢

Magnification behaves multiplicatively across a combination, in contrast to the additive powers. Since the image formed by the first lens becomes the object for the second, the linear magnification of the whole system is the product of the individual magnifications, $m=m_1\,m_2\,m_3\cdots$. Each factor may be positive for an erect image or negative for an inverted one, so the sign of the net magnification records whether the final image is erect or inverted, while its magnitude records the overall size ratio. This product rule, together with the additive power rule, completely characterises a stack of thin lenses in contact. 🔉✕⇢

The importance of lens combinations for optical instruments can hardly be overstated. A compound microscope uses an objective and an eyepiece in series, and its overall magnification is the product of the two individual magnifications, exactly as the product rule prescribes. Telescopes similarly combine an objective of long focal length with an eyepiece of short focal length. In every such instrument the freedom to add powers and multiply magnifications lets the designer reach magnifications far beyond what any single lens of realistic focal length could deliver, which is precisely why multi-lens systems dominate cameras, microscopes and telescopes. 🔉✕⇢

Beyond raw magnification, lens combinations are the primary tool for correcting aberrations. A single thick lens produces coloured images because its refractive index, and hence its power, varies with wavelength, a defect known as chromatic aberration. By cementing together a convex crown-glass lens and a concave flint-glass lens whose dispersions partly cancel, designers build an achromatic combination whose net power is correct while the colour spread is largely removed. Modern microscope objectives and eyepieces are multi-component assemblies for exactly this reason: combining lenses lets one keep the desired total power while minimising the various optical aberrations that degrade image quality. 🔉✕⇢

In problem solving the two rules should be applied with care over signs. When adding powers, convex contributions enter as positive dioptres and concave contributions as negative, so the net power may come out either sign, determining whether the combination converges or diverges. When forming the net magnification, multiply the signed magnifications rather than their magnitudes, so that the erect-or-inverted character of the final image emerges automatically. Finally, remember that the additive power rule assumes the lenses are thin and in contact; when lenses are separated by a finite distance the effective focal length picks up an extra term and the simple sum no longer holds exactly. 🔉✕⇢

Refraction Through a Prism & Minimum Deviation 🔉✕⇢

When a narrow beam of light passes through a triangular prism it does not merely bend once — it is refracted twice and emerges travelling in a decidedly different direction from the one along which it entered. Consider a prism $ABC$ whose two polished refracting faces $AB$ and $AC$ meet along the refracting edge at $A$; the angle $A$ between these two faces is called the refracting angle, or simply the angle of the prism. A ray $PQ$ strikes the first face $AB$ at the point $Q$, making an angle of incidence $i$ with the normal there. Because glass is optically denser than air, the ray bends toward the normal as it enters, travelling inside the prism at an angle of refraction $r_1$. 🔉✕⇢

Full derivation, worked example and interactive 3D on the Refraction Through a Prism & Minimum Deviation tab →

The Simple Microscope (Magnifier) 🔉✕⇢

A simple microscope, also called a magnifier or magnifying glass, is nothing more than a single converging lens of small focal length held close to the eye. Its whole purpose is to let us examine small objects that the unaided eye cannot resolve. A normal relaxed eye can focus an object sharply only when the object lies at or beyond the near point, a distance $D\approx 25$ cm called the least distance of distinct vision. Bringing the object physically closer than $D$ makes it subtend a larger angle, but the eye can no longer converge the strongly diverging rays onto the retina, so the perceived image becomes blurred rather than clearer. 🔉✕⇢

The magnifier removes this limitation. When a converging lens is placed so that the object sits at a distance equal to or slightly less than its focal length $f$, the refracted rays emerge either parallel or only weakly diverging. The eye then interprets them as coming from a large, erect, virtual image located far beyond the object. Because this virtual image lies at or past the near point, it can be viewed comfortably, yet the object itself is held much nearer to the eye than $D$. This is the essential trick: the lens lets the tiny object subtend a much bigger angle at the eye than it ever could unaided. 🔉✕⇢

Two standard viewing configurations are used. In the first, the object is placed just inside the focal point so that the virtual image forms exactly at the near point, $v=-D$. This gives the greatest magnification but forces the eye muscles to accommodate, causing some strain during prolonged observation. In the second, the object is placed precisely at the focus, $u=-f$, so the image recedes to infinity and the eye views it in a fully relaxed state. The relaxed configuration sacrifices a little magnification for far greater viewing comfort, which is why most optical instruments are designed for image formation at infinity. 🔉✕⇢

It is important to distinguish linear magnification from angular magnification. Linear magnification $m=h'/h$ compares the size of the image with the size of the object. Angular magnification, or magnifying power, compares the angle subtended at the eye by the image seen through the lens with the angle the object would subtend if placed unaided at the near point. For a magnifier held close to the eye these two ratios turn out to be numerically equal, but conceptually the angular definition is the honest one, because what governs the apparent size of anything is the angle it subtends on the retina, not its absolute height. 🔉✕⇢

For the near-point setting the working relation is $m=1+\dfrac{D}{f}$. This expression shows that a shorter focal length yields a larger magnification, and that the magnifier always produces $m\gt 1$ since $D/f$ is positive for a converging lens. As a concrete figure, with $D=25$ cm a convex lens of focal length $f=5$ cm delivers $m=1+25/5=6$. To push the magnification higher one must shrink $f$ still further, which is exactly why magnifiers of high power are physically small, thick, strongly curved lenses that must be held very near both the object and the eye. 🔉✕⇢

For the relaxed-eye setting, with the image at infinity, the corresponding result is $m=\dfrac{D}{f}$. This is exactly one unit smaller than the near-point value, a difference that is usually negligible for high powers because $D/f$ greatly exceeds unity. The relaxed formula is the one carried forward when the magnifier serves as the eyepiece of a compound microscope or a telescope, since in those instruments the eyepiece is normally adjusted to throw the final image to infinity for the most comfortable viewing by the observer. 🔉✕⇢

Why does a short focal length translate into high magnifying power? A lens of small $f$ bends the incident light strongly, so it can accept rays from an object held extremely close and still render them parallel or gently diverging. The closer the object can be brought, the larger the angle $h/u$ it subtends, and that angle is what the eye ultimately magnifies. Since $D/f$ scales inversely with $f$, halving the focal length nearly doubles the magnification. This inverse dependence is the single most useful design fact about the simple microscope and about every magnifier built from one converging element. 🔉✕⇢

There are, however, firm practical limits. As $f$ is reduced, the lens surfaces must be ground with ever greater curvature and smaller aperture, and optical defects such as spherical and chromatic aberration grow rapidly, degrading the sharpness of the image. In practice a single-lens magnifier is limited to a magnification of roughly nine or ten before the image quality becomes unacceptable. Beyond that, one abandons the single element and compounds the effect of two lenses, an objective and an eyepiece, which is precisely the reasoning that leads to the compound microscope covered in the next topic. 🔉✕⇢

A subtle but frequently examined point is what actually changes when we look through the magnifier. The angular size of the object and the angular size of its virtual image are, strictly, equal. The instrument still helps because without it the smallest usable object distance is the near point $D$, whereas with it the object may be held at a distance of only $f$, far closer than $D$, so it subtends a much larger angle. The magnifier therefore does not enlarge the object in any absolute sense; it simply lets the eye exploit a viewing distance shorter than the near point while keeping the final image comfortably visible. 🔉✕⇢

The choice between the two magnification formulas is guided entirely by comfort and purpose. When the last ounce of magnification matters, the object is nudged just inside the focus so the virtual image lands at the near point and the eye accommodates to yield $1+D/f$. When the observation is prolonged, the object is set precisely at the focus so the image relaxes to infinity and the eye rests, yielding the slightly smaller $D/f$. The near point $D$ is itself a personal quantity, longer for a far-sighted observer and shorter for a near-sighted one, so the same converging lens delivers a somewhat different magnifying power to different eyes. 🔉✕⇢

In summary, the simple microscope is a converging lens of small focal length that forms an erect, magnified, virtual image of a nearby object. Its magnifying power is $1+D/f$ when the image is set at the near point and $D/f$ when the image is relaxed at infinity, with $D=25$ cm for a normal eye. The magnification rises as the focal length shrinks, but aberrations cap the single-lens instrument near tenfold. These same ideas of angular magnification, near point, focal length, and relaxed versus near-point viewing recur throughout the study of every refracting optical instrument. 🔉✕⇢

The Compound Microscope 🔉✕⇢

A simple microscope built from one converging lens has a limited maximum magnification, typically no more than about nine or ten for realistic focal lengths, because shrinking the focal length further introduces severe aberrations. To achieve much larger magnification one uses two lenses, one compounding the effect of the other. This arrangement is the compound microscope. The lens nearest the object is called the objective and has a very short focal length $f_o$; the lens near the eye is the eyepiece with focal length $f_e$. The objective forms a real, inverted, magnified image, and the eyepiece then acts on that image exactly like a simple magnifier. 🔉✕⇢

The working sequence has two clear stages. Light from a small object placed just beyond the focus of the objective is refracted to form a real, inverted, magnified intermediate image inside the tube. This first image lands at or just within the focal plane of the eyepiece. The eyepiece, functioning as a magnifier, then produces the final image, which is enlarged and virtual, viewed by the eye. Because the objective inverts and the eyepiece preserves orientation, the final image is inverted with respect to the original object, which is acceptable for laboratory specimens that have no preferred up or down. 🔉✕⇢

The distance between the second focal point of the objective and the first focal point of the eyepiece is called the tube length $L$. It is this separation, rather than the object distance, that largely governs how much the objective magnifies. Because the object sits only just beyond $f_o$, the real intermediate image is thrown almost the full tube length away, so its linear magnification is close to $L/f_o$. A large tube length and a very short objective focal length therefore both increase the contribution of the objective to the total magnifying power of the instrument. 🔉✕⇢

The total magnification is the product of the magnification of the objective and the angular magnification of the eyepiece, $m=m_o\,m_e$. This multiplicative rule follows from the general principle that when the image formed by the first lens becomes the object for the second, the overall magnification is the product of the individual magnifications. The objective contributes $m_o=\dfrac{L}{f_o}$, a linear magnification, while the eyepiece contributes an angular magnification $m_e$ whose value depends on whether the final image is set at the near point or at infinity, exactly as for the simple magnifier. 🔉✕⇢

When the final image is formed at the near point of distinct vision, the eyepiece behaves as a near-point magnifier and contributes $m_e=1+\dfrac{D}{f_e}$, with $D\approx 25$ cm. The total magnifying power is then $m=\dfrac{L}{f_o}\left(1+\dfrac{D}{f_e}\right)$. This near-point setting gives the greatest possible magnification for a given pair of lenses, at the cost of some accommodation strain on the eye, because the eye must focus continually on an image held only $25$ cm away throughout the observation. 🔉✕⇢

When the final image is instead formed at infinity, the eyepiece is relaxed and contributes only $m_e=\dfrac{D}{f_e}$, so the total magnification reduces to the compact approximate form $m\approx\dfrac{L}{f_o}\cdot\dfrac{D}{f_e}$. This relaxed expression is the one usually quoted for the compound microscope, because comfortable viewing over long periods demands the image at infinity. It also cleanly separates the two design levers: a short objective focal length and a large tube length maximise $L/f_o$, while a short eyepiece focal length maximises $D/f_e$. 🔉✕⇢

As a worked figure, consider an objective with $f_o=1.0$ cm, an eyepiece with $f_e=2.0$ cm, and a tube length $L=20$ cm, with the final image relaxed at infinity. The magnification is $m=\dfrac{L}{f_o}\cdot\dfrac{D}{f_e}=\dfrac{20}{1.0}\times\dfrac{25}{2.0}=250$. A magnification of two hundred fifty from lenses of centimetre focal length illustrates vividly why the compound design so decisively outperforms any single-lens magnifier, whose ceiling lies near tenfold. Had the same instrument been read at the near point instead, the eyepiece factor would rise from $D/f_e$ to $1+D/f_e$, lifting the total magnification modestly above the relaxed value, which confirms that the near-point setting always yields the larger magnifying power for a given objective and eyepiece. 🔉✕⇢

The expression makes plain why both focal lengths must be small. Since the objective magnification is $L/f_o$ and the eyepiece magnification is $D/f_e$, the total magnification varies inversely with the product $f_o f_e$. Reducing either focal length raises the magnifying power, so a good microscope demands both a short-focus objective and a short-focus eyepiece. In practice it is difficult to grind a lens with focal length much below one centimetre, and making the tube length very large requires correspondingly large lenses, so real instruments balance these competing constraints. 🔉✕⇢

Image quality depends on more than raw magnification. Illumination of the object, the numerical aperture of the objective, and the control of optical aberrations all shape the sharpness and visibility of the final image. For this reason modern microscopes replace each single lens with a multi-component objective and a multi-component eyepiece, groups of elements designed together to minimise spherical and chromatic aberration. A high magnification is worthless if the accompanying defects blur the very detail one is trying to resolve, so professional objectives are corrected assemblies rather than single thin lenses. The useful magnification is also bounded by the resolving power set by the objective aperture; magnifying beyond that limit produces a bigger but no sharper image, a condition known as empty magnification, so the objective is engineered for both a short focal length and a large numerical aperture together. 🔉✕⇢

A practical detail concerns where the object and the eye are placed. The object is positioned only just beyond the focal point of the objective, so that the real intermediate image is thrown far down the tube and magnified strongly by the factor $L/f_o$. At the other end, the eye is not pressed directly against the eyepiece but held a short distance behind it, at the position where all the emerging rays cross, called the eye ring. Placing the eye at this ring lets it collect the maximum light from the whole field of view, giving the brightest and widest final image the instrument can supply. 🔉✕⇢

In summary, the compound microscope cascades a short-focus objective that forms a real, inverted, magnified intermediate image with a short-focus eyepiece that magnifies it further as a virtual image. The total magnifying power is $m=m_o m_e=\dfrac{L}{f_o}\left(1+\dfrac{D}{f_e}\right)$ for the image at the near point and $m\approx\dfrac{L}{f_o}\cdot\dfrac{D}{f_e}$ for the relaxed eye. Both focal lengths must be small and the tube length large to secure high magnification, and multi-element lenses are used to keep aberrations in check while the objective and eyepiece do their compounding work. 🔉✕⇢

The Refracting & Reflecting Telescope 🔉✕⇢

A telescope is designed to provide angular magnification of distant objects rather than of small nearby ones. Like the compound microscope it has an objective and an eyepiece, but the roles are reversed in scale: here the objective has a large focal length and a much larger aperture than the eyepiece. Light from a distant object, arriving as an almost parallel beam, enters the objective and is refracted to form a real, inverted image at the second focal point of the objective, inside the tube. The eyepiece then magnifies this intermediate image, producing the final inverted image that the eye views. 🔉✕⇢

The quantity of interest for a telescope is the magnifying power, defined as the ratio of the angle $\beta$ subtended at the eye by the final image to the angle $\alpha$ that the distant object itself subtends at the objective or the unaided eye. Because a distant object such as a star or a planet cannot be brought closer, its true angular size $\alpha$ is fixed; the telescope's task is to present the eye with a much larger apparent angle $\beta$. This is fundamentally different from the microscope, where the object's distance is ours to choose. 🔉✕⇢

In normal adjustment the telescope is arranged so that the final image is formed at infinity, which is the most comfortable setting for a relaxed eye observing the sky. The real image formed by the objective then falls exactly at the common focal point shared by objective and eyepiece, so the intermediate image sits at the first focal point of the eyepiece. Under this condition the magnifying power takes the clean form $m=\dfrac{f_o}{f_e}$, the ratio of the focal length of the objective to that of the eyepiece. A long-focus objective and a short-focus eyepiece together give high magnification. 🔉✕⇢

The overall length of the telescope tube in normal adjustment is simply $f_o+f_e$, the sum of the two focal lengths, because the intermediate image lies one objective focal length behind the objective and one eyepiece focal length in front of the eyepiece. This is the opposite arrangement to a microscope, whose short focal lengths are separated by a comparatively long tube length $L$. For example, an objective of focal length $100$ cm with an eyepiece of focal length $1$ cm gives a magnifying power of $100$ and a tube roughly $101$ cm long. A pair of stars whose true angular separation is one minute of arc would then appear separated by one hundred minutes, close to one and two-thirds of a degree, which is why even a modest refractor can split double stars that the unaided eye sees as a single point. 🔉✕⇢

Two other considerations dominate the design of an astronomical telescope: its light-gathering power and its resolving power. The light-gathering power depends on the area of the objective, so a larger diameter collects more light and allows fainter objects to be observed. The resolving power, the ability to distinguish two objects lying in very nearly the same direction, also improves with the diameter of the objective. For both reasons the desirable aim is an objective of large aperture. The largest lens objective ever used, at the Yerkes Observatory, has a diameter of about $1.02$ m. 🔉✕⇢

Large refracting objectives are, however, deeply problematic. A big lens is heavy and can be supported only around its rim, so it tends to sag under its own weight and distort the image. It is also difficult and expensive to grind a large lens whose two surfaces are figured accurately enough to be free from chromatic aberration, the coloured fringing caused because a lens brings different wavelengths to focus at slightly different points. These mechanical and optical limits set a practical ceiling on how large a refracting telescope can usefully be built. 🔉✕⇢

For these reasons modern large telescopes use a concave mirror instead of a lens as the objective; such instruments are called reflecting telescopes. A mirror has no chromatic aberration at all, because reflection does not depend on wavelength, so it forms images free of the colour defects that plague large lenses. A mirror also weighs far less than a lens of equivalent optical quality and, crucially, can be supported over its entire back surface rather than only at its rim, so mechanical sagging is far less of a problem and much larger apertures become feasible. 🔉✕⇢

The obvious difficulty with a reflecting telescope is that the concave objective mirror focuses the light back inside the tube, where an eyepiece and observer would obstruct the incoming beam. The Newtonian design solves this by placing a small plane mirror at forty-five degrees to divert the converging light sideways to an eyepiece mounted on the side of the tube. The Cassegrain design instead uses a convex secondary mirror to reflect the light back through a small hole bored in the centre of the objective primary mirror, giving a large effective focal length within a short, compact tube. 🔉✕⇢

These reflecting designs have made the world's giant telescopes possible. The Cassegrain arrangement in particular delivers a long focal length in a physically short instrument, which eases both mounting and housing. The largest telescope in India is a $2.34$ m diameter reflecting Cassegrain telescope at Kavalur, while the largest in the world are the pair of Keck telescopes in Hawaii, each with a reflector ten metres in diameter. No refracting design could reach such apertures, which is why every research-grade astronomical telescope today is a reflector. 🔉✕⇢

Two refinements are worth noting. If the eyepiece is drawn in slightly so the final image forms at the near point instead of infinity, the magnifying power rises to $m=\dfrac{f_o}{f_e}\left(1+\dfrac{f_e}{D}\right)$, a modest increase bought at the cost of eye strain, so normal adjustment at infinity remains the preferred setting for astronomy. Secondly, the astronomical telescope gives an inverted final image, which is harmless for stars but unacceptable for viewing objects on Earth. Terrestrial telescopes therefore insert an extra pair of erecting lenses between objective and eyepiece to turn the final image upright, at the price of a longer tube. 🔉✕⇢

In summary, the astronomical refracting telescope forms a real image of a distant object with a large-aperture, long-focus objective and magnifies it with a short-focus eyepiece, giving magnifying power $m=f_o/f_e$ and tube length $f_o+f_e$ in normal adjustment. Large aperture is sought for light-gathering and resolving power, but the weight and chromatic aberration of big lenses force large instruments to use concave mirrors instead. Reflecting telescopes of the Newtonian and Cassegrain types avoid chromatic aberration, permit huge mirrors supported over their whole back, and so dominate modern astronomy. 🔉✕⇢