ir Foundations of geometrical optics 3.1 Approximation for very short wavelengths ‘Tur electromagnetic field associated withthe propagation of visible light is character- ied by very rapid oscillations (frequencies of the order of 10" s~!) of, what amounts to the same thing, by the smaliness of the wavelength (of order 10° cm), It may therefore be expected that a good frst approximation to the propagation laws in such ‘cases may be obtained by a complete neglect of the finiteness of the wavelength It is found that for many optical problems such a procedure is entirely adequate; in fact, phenomena which can be attributed to departures from this approximate theory (so: called diffraction phenomena, studied in Chapter VIII) can only be demonstrated by ‘means of carefully conducted experiments. ‘The branch of optics which is characterized by the neglect of the wavelength, i. that corresponding to the limiting case 2y — 0, is known as geometrical optics,” since in this approximation the optical laws may be formulate in the language of geometry. The energy may then be regarded as being transported along certain curves (light rays). A physical model of a pencil of rays may be obtained by allowing the light from 4 source of negligible extension to pass through a very small opening in an opaque screen. The light which reaches the space behind the sereen will fl a region the boundary of which (the edge of the pencil) will at frst sight, appear to be sharp. A ‘more careful examination will reveal, however, that the light intensity near the boundary varies rapidly but continuously from darkness in the shadow to lightness in the illuminated region, and that the variation is not monotonic but is ofan oscillatory character, manifested by the appearance of bright and dark bands, called diffraction ffinges. The region in which this rapid variation takes place is only of the order of ‘magnitude of the wavelength. Hence, as long as this magnitude is neglected in ‘comparison with the dimensions ofthe opening, we may speak of a sharply bounded pencil of rays:+ On reducing the sizeof the opening dawn to the dimensions of the * The hire! develo of geometicl apc is deserted ly M. Herzberg, Smalenopi Bet, Spine, 1931) 5.178 Z Psramentenkonde, 2 (1832), 49-45, 48S-W9, 984-5, Cahn, {Geomerscke Opi (Bein, Springer, 1937 and. Mach, The Principe of Phscl Optic 4 steric ‘ant Pasopical Yeament (et German eon 1913, Eaglsh alton: London, Mehae, 1925 ‘ined by DovetPoblcaons, New Yor 1553 1 Tht the Roundy becomes hap in he mit a y ~ O was fi shown by G. Kits Jotenangn 3 Math Phy, Vol 2 (Matematsche Opi) (Lepi,Tuetne, 891}, p13 See ale BB Baler and ET {Capon The Maihematal Theory of Huygens Pinole (Oto, Caecon Pr, od eon, 1950), [Link] A Sormertla Opies (Now York Academic Pro, 195), $5 621 Approximation for very short wavelengths uw wavelength phenomena appear which need more refined study, If, however, one considers only the limiting case of negligible wavelengths, no restriction onthe size of the opening is imposed, and we may say that an opening of vanishingly small ‘dimensions defines an infinitely thin peneil — the light ray It will be shown that the Variation inthe cross-section of a pencil of rays is a measure of the variation of the intensity of the light. Moreover, it willbe seen that iis possible to associate a state of polarization with each ray, an to study it variation along the ray. Further it will be seen that for small wavelengths the field has the same general character as that of a plane wave, and, moreover, that within the approximation of {geometrical optics the laws of refraction and reflection established for plane waves incident upon a plane boundary remain valid under more general conditions. Hence if «light ray falls on a sharp boundary (€-. the surface ofa len) its split nto a reflected ray and a transmitted ray, and the changes in the state of polarization as well a8 the ‘reflectivity and transmissivity may be calculated from the corresponding formulae for plane waves. ‘The preceding remarks imply that, when the wavelength is small enough, the sum total of optical phenomena may be deduced fom geometrical considerations, by determining the paths of the light riys and calculating the associated intensity and polarization. We shall now formulate the appropriate laws by considering the implica- tions of Maxwell’ equations when 2 —>0.* BLT Derivation of the eikonal equation ‘We consider a general time-harmonic field tr, 1) = Fo(r}e“™, } H(t, 1) = Hole", in a nonconducting isotropic medium, Eo and Hy denote complex vector functions of positions, and, as explained in §1.4.3, the real parts of the expressions on the right hand side of (1) are understood to represent the fields, ‘The complex vectors Ey and Hy will satisfy Maxwell's equations in their time-free form, obtained on substituting (1) into §1.1 (1)=(4). In regions free of currents and ges (j = p = 0), these equations are o curl Hy + koe = 0, @ curl Eg ~ itl = 0, @ diveEo =0, ® iv uly = 0, © "ast shown by A. Sommerd an Runge Poe 3 (191), 29, sing a suggestion of Debye tate base egeaton of geome ps te eka union (38) oy be dee om the (cai) wave equation nthe img cave yO Generale which tke ine focus Share ofthe lesromgete Nel ate det W. ents, Dane State Ope Iti Pray 40918), No 3.30, Tock, th 3.0924), 318. MER, Comp Rend (Dotad) dead Se URS 18 (938), 26. N Ay Det. Kel Dont Vine Ses. 38 (1345, No. FG Fede, Poe amb: Pi Sec 48 (197), 284K. Seay dw Pha 14 (1952), 11, td (1953 983d id, 131955, 178 R-'.Ingtéen and A: Key, cn Phy, Plot, 41989, 258,us I Foundations of geometrical optics Here the material relations D= cE, B= WH have been used and, a8 before, fy = w/e = 2, o being the vacuum wavelength ‘We have seen that « homogeneous plane wave in «medium of refractive index ‘n= J&R, propagated in the direction specified by the unit vectors is represented by Fyaeshrr, Hy a heltonen, 6 ‘where ¢ and h are constant, generally complex vectors. Fora (monochromatic) electric pole field in the vacuum we found (see $2.2) that By =e, Hy hole, o +r being the distance from the dipole, Here ¢ and h are no longer constant vectors, but at distances sufficiently far away from the dipole (r >> 4g) these vectors are, with suitable normalization ofthe dipole moment, independent of ky ‘These examples suggest that in regions which ate maay wavelengths away feom the sources we may represent more general types of fields in the form Ey= aryl, ay ribs, © where S(t), ‘the optical path’, isa real scalar function of postion, and e(r) and b(r) are vector functions of positon, which may in general be complex." With (8) as atrial solution, Maxwell’ equations lead toa set of relations between e, hand S. It will be shovin that for large fy (small 2o) these relations domand that S should satisfy @ certain differential equation, which is independent ofthe amplitude vectors e and h. From (8), using well-known vector identities, curl Hp = (curt + iky grad S x mye, o div lg = (divin +h grade + ikpgth - grad Se, (10) with similar expressions for curl Ea and div Fp. Hence (2)-(S) become ands Xhpc0=— ay grid x e~ ph = - «a e-gnds = — 7 (e-gndine + dive), (3) he grad = qt arading + dv 4) We are interested in the solution for very large values of ko, Hence as long as the ‘multiplicative factors of 1/i on the right-hand side are not exceptionally large they ‘may be neglected, and the equation then reduce to * Complex ean ware neces if al one ss of olariaion ae to be included, Accoding to 1.4 {7s} teal eandhcocetpond i fds wich tinal pole,31 Approximation fr very shot wavelengths nu grads x h-+e¢ (a) grdS X e— ph = (12) e-grds (13a) hegrad S =0. (148) ‘We can confine our atention to (I1a) and (12a) alone, since (13a) and (14a) follow from them on scalar multiplication with grad S, Now (113) and (2a) may be regarded as a set of six simultaneous homogencous linear sealar equations for the Cartesian components ¢, tay -.. of € and h These simultaneous equations have nontrivial solutions only if @ consistency condition (the vanishing of the associated determinant) is satisfied. This condition may be obiained simply by eliminating e or h between (1a) and (12a). Substituting for h from (12a), (11a) becomes Fle endS) ends ~ eignds'] + 4e=0. ‘The first term vanishes on account of (13a), and the equation then reduces, since does not vanish everywhere, to (15a) (ends? of, writen explicitly, TG +@l-ra om ‘where as before n= YEH denotes the refractive index. The function S is often called the cikonal* and (15b) is known as the eikonal equation; it is the basic equation of ‘geometrical optcs.t The surfaces se) ‘may be called the geometrical wave surfaces or the geometrical wave-fronts The eikonal equation was derived here by using the fs-order Maxwell’ equations, but it may also be derived from the second-order wave equations forthe electric or ‘magnetic field vectors. To show this one substitutes from (1) and (8) into the wave equation §1.2 (5) and obtains, after a straightforward caleulation, stant KOS +E Ue Smt aeeMe 60) a6) Teo where * The term cdonal tam Geek eon = image) was introduced in 195 by H. Brust dsr sein ‘ated into (0 p42, buthas came to be ed ins wer one + Theekonal equation may also ered the eguton of he characters ofthe wave eatin 1.2 (Sy and $1.2 (6) Tor Eat I ana cescbes the propagation of dacontatis ofthe slate of tae ‘ution In ponent optes we a, however at Coszened wth the propapton of dco ut ‘thine ton ory time uronic) slurs Te fra auc of et eens Irdemonarte in ApgenitV ‘Te eikonal equation wil ako be recognized as the Hariton-acoi equation ofthe variation problem dfn he opal orp of which ges back to Fema (oe $33.24 Append, 1 Infuure we stall op the adjective “geomet when tena ak of conto,120 It Foundations of geometrical optics K(e, 5, m) = (0? ~ (grads), Lee, S, n, w) = [grad S- grading ~ V*S]e — 2[e- gradin m] grad S = U{grad 5 - grade, ‘Mee, €, 1) = curle X pradin ~ Ve — grad(e- gradine) ‘The corresponding equation involving h, obtained on substitution into the wave equation §1.2 (6) for H (or more simply by using the fact that Maxwell’: equations remain unchanged when E and H and simultaneously ¢ and — are interchanged), is 1 Toe For sufficiently large & the second and third terms may in general be neglected; then K = 0, giving again the eikonal equation. It wil be seen later thatthe terms in ‘the frst power of Ii in (16) and (17) also possess a physical interpretation {tmay be shown that in many cases of importance the spatial parts Fe and Hy ofthe field vectors may be developed into asymptote series of the form* KO, 5,0) + US. nO eM, W.0)=0. an x Ji08 19) Sie “ where e and h( ate functions of positon, and S is the same function as before:t Geometrical optics corresponds tothe leading terms ofthese expansions. 4.1.2 The light rays and the intensity law of geometrical optics From (8), and from §1.4 (S4) and $1.4 (55), it follows thatthe time averages of the lectic and magnetic energy densities (w.) and (14) are given by =ahere, me) = einen (rd = Teer’, om) = fem « ‘Substitution for e* from (118) and for from (12a) gives ee Fel bared}, 20) the square bracket denoting the scalar triple product. Hence, within she accuracy of geometrical optics, the time-averaged electric and magnetic energy densities are ‘equal. ‘The time average of the Poynting vector is obtained from (8) and §1.4 (56) (8) = EReXb), (7) = (00a) = We assume here tat caly cee geometrical wav-oot pastes though ech point, In some cases, for ‘example whe redeedn tats pce st cbsacles preset i he meduny seve wavefronts may Pst "heough enc oi The ebling Sl en epee bythe atom af are te sve pe 4 The theory of ach asymp expnaons far engin chef mth werk of RK Coeba. Propagation of Elecromogntc ove ancora! lars tee New York Unver, 1947" 1588. See alsoM, Kine, Comm. Pie nd App Math. 4(1951) 235 i 8 (193), 599 and W. Bre 2 Nanerch, 6 (1951), 672.4 compreenive account ofthe theory it ela M. Kiet 1 Way, Electromagnetic Thor and Geomorel Opes New Yer, fensene abs, 1963).3.1 Approximation for very shor wavelengths ma Using (12a), we obtain = Hg lorena — (e+ grad S)e*}. ‘The lst term vanishes on account of (138) so that we have, if seis made of the expression for (w,) and of Maxwell’ relation ey. (8) = 7500.) ends. ey Since (w.) = (ws the term 2(;) represents the time average () ofthe total energy density (i.e. (w) = (we) + (Win). Also, on account of the eikonal equation, (grad S)/m is a unit vector (s say), gnds_ grads = Tgradsi* o and (21) shows that sis inthe direction of the average Poynting vector. If, as before, wwe set c/n = v, (21) becomes 8) Hence the average Poynting vector isin the direction ofthe normal to the geometrical wavefront, and its magnitude is equal 10 the product of the average energy density and the velocity v= c/n This result is analogous to the relation §1.4 (8) for plane waves, and shows that within the accuracy of geometrical optics the average energy densityis propagated wit the velocity v = cj The geometrical light rays may now be defined as the orthogonal trajectories to the ‘geometrical wave-ftonts S = constant. We shall regard them as oriented curves whose direction coincides everywhere withthe direction of the average Poynting vector.” If (3) denotes the position vector of a point P on a ray, considered as a function of the length of are sof the ray, then dr/ds = s, and the equation of the ray may be written as ws @) ” 4) From (13a) and (14a) itis seen that she electric and magnetic vectors are at every ‘point orthogonal tothe ray. ‘The meaning of (24) may be made clearer fiom the following remarks. Consider two neighbouring wavesfronts S = constant and $ + 4S = constant (Fig, 3.1). Then as_ G& Hence the distance ds between points on the opposite ends of a normal cutting the two ‘wave-fronts is inversely proportional to the refractive index, i. directly proportional toe. The integral {cds taken along a curve C is known as the optical length of the erdS =n, 2s) * This desnion or igh ms is appropiate a isotope media ony, Weal eer (Chapter XV) hat a tn ansoapic metum the Seton ofthe wave our norma dos hin Beer, emi wih he Seton of the Ponting vector