252 ON THE THEOKY OF THE MICROSCOPE. 



mentioned in paragraph x,, where an artificial test object is 

 illuminated with two isolated pencils of light. The divergence of 

 the first difi'raction pencil obtains in this case such a relation to 

 the angular aperture of the objective ^that — as theory and direct 

 observation of the tracks of light shew — by setting the mirror in 

 two special directions, parts of all the zones of the aperture, each 

 represented by separate lines of light, will be set at work under 

 circumstances which strongly conduce to bring out any existing 

 failure of correction. One position of the mirror would be, when 

 placed with its inner edge just outside the axis of the instrument, 

 the whole of the mirror being then on one side of the axis and its 

 surface turned at right angles to the striae in the object, so that 

 the track of the t/w'^c^J incident light would appear in the aperture- 

 image above tlie objective, close to its centre, whilst the track of 

 diffracted ray would appear in the periphery on the opposite side. 

 The second position would be that of greatest oblique illumina- 

 tion which the objective would bear without marked loss of light. 

 The tracks of the two pencils would, as soon as this change of 

 illumination was effected, simply exchange places. In both cases 

 there would be — supposing the object to contain only one set of 

 striae — two isolated pencils set in action which would engage a 

 portion of the central and a portion of the peripheral zone of the 

 apertuie on opposite sides of the axis at the same time. But if 

 the object contained several uniform sets of striae, although 

 additional diffraction rays would pass through the objective, no 

 essential change of the relations before[noted would take place. 



The intention of this procedure is not, of course, to discover 

 each particular fault in the image-forming quality of an objective, 

 as can, indeed, be done by the method described in paragraph 

 (x ), but to test, in a general way, the actual performance of a 

 lens in such a manner as would represent the normal state in 

 ordinary use of the microscope. The factor of chief practical 

 importance reveals itself at once when attention is given to the 

 degree of correctness with which the fusion of the several partial 

 images which belong to one and the same part of the object takes 



