Appendix 125 



central beam wherewith to form an image of the grating, 

 the image would be almost identical with that which would 

 be furnished by light coming through an opening covered 

 by tinted glass, and no trace of the ruling would be seen 

 in it. In order to see the ruling, the telescope lens 

 must be able to catch and forward to its focus other 

 rays which have passed through the grating than those of 

 the central beam. The more of the lateral beams which it 

 can transmit and combine at its focus with the light of the 

 central beam (where they will by interference strengthen 

 some parts of the image formed by the central beam, and 

 enfeeble others, thus introducing detail), the more nearly 

 will the image it forms resemble the actual grating in detail, 

 and in freedom from false colour. If it succeeds in catching, 

 along with the central beam, even some small portion of the 

 nearest of the diffraction beams, the image will exhibit lines, 

 and the proper number of lines, though it will not present 

 correctly such minuter detail as the widths of the lines and 

 of the spaces between them. 



Cases exactly analogous to this occur with the micro- 

 scope. When an object covered with dots, such as Pleuro- 

 sigma angulatum, has been focused upon the stage, and 

 is resolved, the diffraction beams may be clearly seen upon 

 the back lens of the objective by removing the eyepiece 

 and looking down the tube. With this diatom there will 

 then be seen the central beam, and portions of the nearest 

 of the lateral beams, six in number. A rather small cone 

 of illumination is best to show them conspicuously if white 

 light be used, and they can be seen with larger cones of 

 illumination and very sharply denned if monochromatic 

 green light, produced by prisms, be employed. 



The markings on the Pleurosigma angulatum are spaced 

 in each row at intervals which have been measured, and 

 found to be equal to wave-lengths of red light. With so 

 close a ruling the lateral beams are much diffracted or bent 

 aside, and dry objectives can only take in the central beam 

 and a portion of each of the nearest diffracted beams. This 



