W. K. BRIDGJIAN ON A NEW REFLECTING ILLUMINATOR. 215 



of certain definite parts of the spectrum, and consist of rays 

 approaching somewhat to parallelism, or be in one plane, and not 

 convergent upon the object from all points of a circle ; next, it is in- 

 dispensable that the light should impinge upon the object at that 

 particular angle which serves best to bring out its minutest detail 

 with the greatest sharpness. 



With perfectly transparent material, such as the siliceous frustule 

 of a diatom, it is only at some particular angle, to be found by 

 experience and patience, that the so-called markings are displayed. 

 Now, with a flint-glass prism, light undergoes total reflection from 

 its interior at an angle of 41° 48", and, therefore, with the flint 

 frustule of the diatom, its net-like structure causes it to present 

 surfaces at the angle producing this total reflection, and hence, 

 where the light is stopped, the lines, bars, or dots appear dark, or 

 as shadows by contrast with those parts through which the light i*s 

 transmitted. When the light is, however, merely refracted unequally 

 by inequality of thickness, or by varied direction of surface, the lines, 

 dots, or shadows are only faint or dark in proportion to the amount 

 of refraction produced, but not sharp and deep like those arising 

 from total reflection ; and hence arises the extreme nicety of adjust- 

 ment necessary to " hit off" the exact position of the frustule and 

 the precise direction of the light required to obtain the greatest 

 effect. From this will be understood the necessity of stopping 

 out all cross rays, or rays coming in opposite directions, and which 

 principle applies equally to every description of transparent 

 material. 



With the prism, or with the concave mirror, we can obtain a 

 degree of obliquity only up to a certain point, when it becomes 

 restricted by the thickness of the stage ; but if we direct a cone of 

 rays from a small condensing lens, placed in the position shown at 

 L, in Fig. 1, so as to throw its focus beyond the centre and to the 

 point A, and then interpose a small metallic speculum S vertical to 

 the plane of the stage, the intercepted portion of the light will 

 hence be reflected back to a focus at B — a point in front of the 

 reflector equidistant with the point A beyond it, and it will thus be 

 made to reach the object irrespective of any amount of st&ge-thick- 

 ness. It will now be seen by the different positions of the reflector 

 S' S" 8'", in Fig. 2, that to obtain light at any particular angle 

 from its greatest possible obliquity down to direct rays, it will only 

 require a very slight alteration in the position and the angle of the 



