LIGHT. 244 LIGHT. point-sources — that is, to sec them as two eep- urate objects — if thej- arc as close to{;ettier as a certain traction of the wave-lcngtli of the liglit which they arc eniilting. This fact depends upon the properties of dillraetion of the ether-waves, and will be discussed later. For detailed de- scription of various optical instruments, see the articles Microscope; Telescope; Spectro- scope, etc. PHYSICAL optics. In this branch of light the assumption is made that the sensation called light is due to the re- ception into the eye of trains of waves. A 'source of liglif is then a source of waves, which are propagated from the source to the eye — or to whatever instrument is used to detect their pres- ence: — and produce there certain ell'ects. The phenomena to be considered are then: (1) the fact that trains of waves carry energy: (2) the fact that waves are propagated with a finite velocity; (3) the kinematic properties of com- position of wave-motion. PiioTO.METRY (q.v.) is that division of 'light' which is concerned directly with the comparison of quantities of light emitted by various sources and received by dillerent bodies. If a point- source emits in all directions a 'quantity of light' M, its 'intensity' is said to be M/4ir. i.e. the quantity of light going out through a unit solid angle. This intensity is written I. A small sur- face of area A, at a distance r from the point- source, and inclined to the line joining it to the point-source so that the angle l)etwecn this line and a line perpendicular to the surface is S, re- I.cos9 ceives therefore an amount of light ^ — . The intensity of 'illumination' of this surface is the amount of light per square centimeter ; it is Ico«9 written E. Hence E = — 5 — . Therefore the illumination varies directly as the intensity of the light, inversely as the square of the dis- tance, and as the cosine of the obliquity of the surface. Velocity of Light. The fact that time is re- quired for the propagation of whatever it is that causes light lias been known since the observation of Koemer on the eclipses of .Tupiter's .satellites in 1()7.5. It follows then — granting that light is due to wave-motion — that there is a medium tilling all interstellar space, which serves to convey these waves. This medium is called 'the ether' (q.v.) , and it is evident from the prop- erties of transparent bodies that it permeates them, and may in fact be regarded as a tmiversal medium. There is every reason, both theoretical and experimental, for believing that the velocity of waves of all lengths is the same in the pure ether, e.g. in interstellar space: but inside of ordinary matter the vclncity of ether-waves varies with the wave-length, and of course their velocity is dilTcrent in difl'crent media. It will be shown below under Refi-iidion that the index of refrac- tion of one medium with reference to another for a definite color is the ratio of the velocity in the second medium of these ether-waves which correspond to that color to that of the same waves in the first medium. There are two experimental methods for the determination of the velocity of ether-waves in air — or. as often expressed, the 'velocity of light' in air. One depends upon the use of a toothed wheel, which is made to revolve rapidly in front of a source of light, thus al- lowing iiitcrniitlcnt Hashes of light to be seen through the teeth. These waves coming through the gaps between the teeth traverse a considerable distance — .several miles — fall upon a mirror which reilects them back in their original direc- tion. If on their return to the wheel the latter has turned so far that they strike a tooth, they are stopped; if, however, the wheel has turned farther .so that they find an opening between the teeth, they will pass through and their return may be observed by suitable means. If the speed of the wheel is thus exactly r-sht, the waves |)ass- ing out through one opening will return through the next; if the speed is now increased, the re- turning waves will be slopped by a tooth; if the speed is still further increased, the waves passing out through any one opening will return through the next but one, etc. Therefore, if the speed of the wheel, the size of the teeth, and the distance between the wheel and the mirror arc known, the velocity of the waves may lie calculated. This method is due to Fizeau and has been u.sed by him, by Cornu, by Forbes an<l Young, and more recently by Perrotin. The other method consists in allowing light from a narrow slit to fall upon a mirror, lie reflected to another distant mirror which retlects it back to the first mirror and thus back to the source. This first mirror is not sta- tionary, however, but is made to revolve rapidly: so that, when the waves return from the distant mirror, the angles of incidence and relied ion are not quite what they were before, and so the waves are not reflected directly back to the slit-source, but are deflected slightly. If the amount of this deflection, the distance apart of the two mirrors, and the rate of revolution of the rotary mirror are known, the velocity of the waves may be cal- culated. This method is due in part to Fizeau also, but in the main to Foucanlt. It has been used by the latter, by Michclson, and by New- comb. For several reas(ms this method is not as good as that of the toothed wheel. (See Cornu. Ifcports of International Congress of Phi/sics, Paris, 1000. vol. ii., p. 22.5.) The accepted value for the velocity of light in air is 300,000 kilometers or 3 X 10" em. per second, with a possible error of less than one part in 1000. This is about 186,- fiOO miles per second. Foucanlt showed by direct experiment that the velocity of light in air is greater than that in water, an observation which definitely overthrows Newton's corpuscular theory of light. Huygexs's Principle. To show how the prop- erties of rays, which were assumed as the basis of geometrical optics, are conseqtiences of the propagation of ether-waves, use must be made of a theorem in kinematics known as 'Hnygens'a principle.' As a result of disturbances in any medium which can transmit wave-motion there will be at any instant a definite wave-front — that is. a surface which is the locus of all points just reached by the spreading waves : there — will be no efl"ect at any point ahead of the wave- ■ front until it is reached : but Huygens's principle ■ asserts that we can predict what will be the effect at that point if we assume that each minute por- tion of the medium in the wave-front is a centre of spherical waves. These spherical waves spreading out from each point of the wave-front all start in the same phase; but. as they overlap and are superimposed at the point for which the efTect is to be predicted, they will be in different
