112 W. B. STOKES ON IMAGES OF DIATOM STRUCTURE. 



is to say, they behave as do the images of ordinary diatomic 

 perforations. 



Pleurosigma formosum is either optically or structurally 

 pecuUar. I cannot imagine the former, so I suggest the latter. 



The possibility of the former would raise doubts as to whether 

 the microscope were at all a reliable instrument ; but my 

 suggestion that P. formosum has its perforations silted up 

 explains all former difficulties in the interpretation of the images 

 of this diatom, and, besides, makes the excef)tion confirm a 

 method for the interpretation of such images as we are con- 

 sidering. And what is that method ? It is a comparison of the 

 relative positions and appearances of the images of unknown 

 structures with those of the images of known structures. 



For instance, assume the under-corrected objective and the 

 axial cone, and we notice that clear diatomic perforations, 

 fractures, grooves, etc., are bright at the uj^per focus and dark 

 at the lower focus ; while the silex, ribs and thickenings of 

 silex especially, are dark at the upper focus and bright at the 

 lower focus. You may have a peculiar line on a siliceous valve 

 to interpret. According to the rule, if it be bright above and 

 dark below, it is a fracture, groove, or tube ; if it be dark above 

 and bright below, it is a rib or other increase in thickness. I 

 believe this rule to be applicable to all objects examined under 

 these conditions. 



But which, after all, is the true image ? The answer must 

 await a definition of a true image. If an image be only true 

 when it has some macroscopic analogy, then only the " white 

 dot " image is true. If, however, an image be true when it leads 

 to a truthful interpretation of structure, then the "black dot" 

 image is also true. Both images, in fact, agree in contour and 

 detail with the object — that is, as far as aperture diffraction 

 effects will permit. 



Jov,m. Quekett Microscopical Club, Ser. 2, Vol. VIII., No. 49, November 1901. 



