Br. G. J. Stoney on Microscopic Vision. 435 



from all the directions in which the iris diaphragm and stops 

 allow light to pass, and at the same time in all other respects 

 — in phase, ellipticity, and so on — such light as would have 

 been emitted in the reverse direction downwards from a 

 perfectly featureless self-luminous plane occupying the position 

 upon the stage of the microscope of the objective field. 



This ideal is the more nearly approximated to, the better the 

 condenser ; and it would appear that the best position of the 

 source of light is that which would be occupied by the image 

 which the condenser would form beneath of such a featureless 

 luminous plane as we have supposed. Hence the suggestion 

 as to the position of the source of light made in § 21. It 

 further appears that the ideal position for the iris diaphragm 

 and stops would be at the position which we may call z 

 (corresponding to x and y) where beams of uniform plane 

 waves emitted downwards from the supposed luminous plane 

 would be brought to a focus by the condenser. This is a 

 position which is usually very close to the condenser ; and it 

 would be a marked improvement in microscopes if the iris 

 diaphragm and stops were brought nearer to this ideal position 

 than they commonly are. As they are at present placed, 

 different parts of the field of view are treated differently by 

 them in an appreciable degree. 



We have next to examine into some points the consideration 

 of which will put us in a better position for interpreting 

 aright what we shall see in the microscope. 



28. Of the Composition and Resolution of Undulations. — We 

 shall start from the known fact that any luminous undulation 

 of uniform plane waves of wave-length X may be resolved 

 into two undulations of plane-polarized light polarized in 

 planes at right angles. In order to get their equations in 

 their simplest form, let the axis of x be placed perpendicular 

 to the wave surfaces, and the axes of y and z parallel to the 

 transversals of the two plane-polarized components. Then 

 the equations of the given undulation, which we may call U, 

 will be — 



— (vt — x)+ct\, . . . . (1) 



£=b sin/ 



*s-(rtJ # ) + A .... (2) 



in which rj and f are the displacements, at the time t, in the 

 two plane-polarized components, a and h are the amplitudes 

 (i. e. the scalar part of the transversals), v is the velocity of 

 light, and a and /3 are the initial phases on the plane y:, 



