Theories of Microscopical Vision. By A. E. Conrady. 633 



themselves, and, therefore, to the law of phase which I have 

 originally deduced for plane waves. 



The whole reasoning of my paper may, therefore, be applied to 

 this case of spherical wave -fronts. The latter will be refracted by 

 the objective according to their apparent point of origin, which, in 

 the case of fig. 101, is a little below the plane of the grating. The 

 only difference as compared with figs. 97 and 98, will, therefore, be 

 that the conjugate foci P and P' will lie much closer to the plane 

 1 1 of the true image, and that the space within which spurious 

 interferences can take place is correspondingly curtailed. But as 

 we can always limit the recognisable image to the plane 1 1 by 

 using an extended source of light, there is no very evident advan- 

 tage in this one difference in favour of spherical wave-fronts. 



I hope I have now made it clear that it must always be very 

 risky to assume that an object might be treated as self-luminous ; 

 that we must, therefore, seek the explanation of microscopical 

 images on the basis of the great principle laid down and proved 

 by Prof. Abbe, and that this latter principle, when applied to ordi- 

 nary working conditions, leads to the formation of images of plane 

 gratings which are as close replicas of the original as could possibly 

 be desired. 



In a supplementary paper I hope to apply the Abbe theory to 

 crossed lines and to dot-patterns. I shall then be in a better 

 position to discuss the question of favourable illumination, and to 

 show what is the real cause of the pleasing images obtained by 

 the use of carefully regulated " aplanatic cones " of light. 



