TESTING THE CHROMATIC ABERRATION. 161 



Jiuiieche's Xos. 7 and 9 objectives. Others are achromatic for 

 the marginal parts of the field, though under-corrected for the 

 central. 



We have tacitly assumed that the coloured borders, which arise 

 from the chromatic aberration of the objective, are seen by the 

 eye unaltered. This supposition was admissible for a preliminary 

 consideration where the question was chiefly relating to qualitative 

 and not quantitative differences in the phenomena ; but it is not, 

 however, in any case strictly accurate, inasmuch as the refractions 

 in the eye-piece, and in the eye of the observer, are accompanied 

 by colour-dispersion. These points merit an exhaustive discussion. 



As regards the eye-piece, it is a widely-spread error, which 

 has become stereotyped in micrographic works and text-books of 

 Physics, that the combined action of the field-lens and eye-lens 

 compensates, to some extent, the defective achromatism of the 

 objective. According to the usual representation we are directed 

 to construct, by means of the rays of direction, the red and violet 

 image produced by the objective, as well as the corresponding 

 images produced by the field-lens and eye-lens (i.e., of the rays 

 which pass through the optical centre of the lenses), in which case 

 it is, of course, always possible to give to the eye-lens such a posi- 

 tion that the final virtual images are seen almost at the same 

 angle, and must therefore coincide. This theory of the Campani 

 eye-piece, as developed, for instance, in Mueller - Pouillet's 

 " Physik," Harting's " Das Mikroskop," &c., is entirely erroneous, 

 as will now be shown. The conclusions are accurate, but the hypo- 

 theses are wrong. 



It has already been shown that the real image formed by the 

 field-lens appears under altogether peculiar conditions, in con- 

 sequence of which its construction cannot be found by simple 

 drawing of the rays of direction. This obtains likewise for the 

 virtual image of the eye-lens. Eacli point of these images is 

 formed by a surface-element of the lens, different from the adja- 

 cent one. The further the image-point is from the centre of the 

 field, the more remote also from the centre of the lens is the 

 active part of the refracting surface. Moreover, the incident cones 

 of light are so reduced where they meet the lens that they are 

 hardly 1 mm. in diameter, and are therefore refracted almost as 

 simple rays. 



Having established the actual relations, which are ignored in 



M 



