250 ON THE THEORY OF THE MICROSCOPE. 



but whicli need not be repeated for any other kind of operation ; 

 for detail only approaches the limit of resolving power when it is 

 so minute, and causes so strong a dispersion of light by diffraction, 

 that even under the most favourable circumstances only the first 

 deflected pencil can enter the objective simultaneously with the 

 direct rays. And when it is visible in the image, this is only 

 accomplished by the operation of the outermost j^eripheral zone of 

 the aperture. The most oblique incident pencil which the mirror 

 can deliver streaks the edge of the aperture on one side, and the 

 single diffracted pencil which gains access streaks it on the other 

 side, as may be proved by direct observation of the tracks of both 

 pencils in the upper focal plane of the objective. But theory and 

 practice teach us that every objective which is not a total failure 

 — however imperfect in respect to correction of spherical aberra- 

 tion — if its lenses be but moderately well centred, can always be 

 made to work with one of its zones, e.g., the outermost, and this 

 permanently, if during its construction it has -been tried on a 

 similar test; and if it be furnished with a correction collar 

 (adjusted during use, to throw the action upon the periphery of 

 the lens), an arrangement which is, in fact, much oftener 

 employed for this than for its ostensible 'purpose (correction, 

 namely, for thickness of glass cover). 



The proof that an objective can resolve very minute strice on a 

 diatom or I^obert's test plate, attests, strictly speaking, nothing 

 more than that its angular aperture answers to the calculable 

 angle of diffraction of the interlinear distance of the striae on the 

 test, and that it is not so badly constructed that a sufficient 

 correction of its outer zone is impossible. A. trial of this sort 

 offers no means of ascertaining what conditions for the correct 

 fusion of aperture images such an objective would present in the 

 much more unfavourable case of the ordinary observing position, 

 where one or more zones in various parts of the aperture 

 are almost always taken up at one and the same time with 

 rays that are in effective operation. Nor can the result be 

 considered as sufficiently characteristic even of the ''resolving 



