546 



Transactions of the Society. 



former paper, we may safely assume that all the spectra imme- 

 diately surrounding the direct light will leave the centres of the 

 dots in exact phase with the direct light, and will, in accordance 

 with the fundamental principle of the equality of optical paths, 

 arrive in the same phase-relation at the centres of the ideal geo- 

 metrical images of the dots. Hence we see that in this case also, 

 in precise analogy to what I proved to be the case with simple 

 line-gratings, the centres of the bright dots must be represented in 

 the image by maxima of brightness exactly coinciding with the 

 geometrical images of the dots, and that the position of the latter 

 will, therefore, be correctly indicated. 



Fig. 127a. 



Fig. 127b. 



Fig. 127c. 



Fig. 127b. 



Fig. 127b. 



It will be highly instructive to study the image in its gradual 

 evolution as the aperture of the Microscope object-glass is in- 

 creased. 



In order that any structure may be shown at all, we must have 

 at least two maxima entering the objective — say the direct light and 

 one of the innermost spectra. If we admit the direct light and the 

 spectrum Aj in the manner illustrated in fig. 127a, the Microscope 

 will show the lines a in fig. 125, from which the spectra A x , A 2 , 

 etc., are derived. If we admit the direct light and the spectrum 

 Bj as shown in fig. 127b, we obtain an image displaying the corre- 

 sponding lines b of fig. 125. Evidently, either of these images is 



