

-. \k 



TJNIVEB 



OBLIQUE ILLUMINATION. 251 



a decided preponderance, that combination is always the most 

 favourable by which the shadow lines and the bright lines are of 

 nearly equal breadth, and the former, moreover, reach their 

 greatest depth. Where both these conditions are not attain- 

 able simultaneously, a greater breadth of shadow can be produced 

 only at the cost of its intensity, and a greater intensity only at 

 the cost of its breadth. The eye must then decide what medium 

 point is the most serviceable for observation. 



As regards oblique illumination by the marginal rays of an 

 axial cone of light of tolerably large aperture, such as may be 

 obtained by the application of an annular diaphragm, we have 

 not had the opportunity of becoming practically acquainted with 

 its advantages in any of our researches. As far, however, as 

 experiments made with test-objects admit of our judging, we 

 should not value them very highly. Theoretically considered, 

 the effect of this kind of illumination, as far as regards the 

 dioptric part of the question, can only be to decrease the intensity 

 of light in the image in the same ratio as that in which the 

 transverse section of the effective cone of light diminishes through 

 shutting off the central rays. The bounding lines of the umbra 

 and penumbra retain their position under all circumstances, as 

 they are in every case dependent only upon the inclination of 

 the extreme marginal rays i.e., upon the maximum values of 

 S and a). The intensity of light may also be reduced by directing 

 the mirror to a less intense (yet sufficiently extensive) source of 

 light, or by using a surface of white paper as a reflector, or by 

 projecting the incident rays through a semi-transparent substance 

 (tissue paper, ground glass, &c.). 



An essentially favourable action of this method of illumination 

 would be intelligible if the objective were aplanatic for marginal 

 rays within certain limits of inclination only, but not for central 

 rays, or if the peripheral interference pencils acted more favour- 

 ably alone than in combination with the more central ones. In 

 the former case an annular diaphragm would be advantageous 

 not merely for particular objects, but for all without exception. 



