8 THE AMERICAN MONTHLY [Jan 



as many diameters. He further explains his work IThus : 



I use the best known form of microscope and prepare 

 the slides and slicings and stainings in the usual way ; 

 and focus and illuminate so as to get the clearest and 

 highest magnification of the object, when viewed through 

 the usual ocular. Then I remove the outer lens of the 

 ocular. It can be shown that the "virtual" image pro- 

 duced by the ocular and eye, although it looks much 

 larger than the "real" image, adds no new details to the 

 real image. This fact is known to many modern micro- 

 scopists. I therefore use the "real" image as the starting 

 point for my new microscope. 



I bring down upon this "real" image or "focal plane" 

 the objective of my second microscope, and thus magnify 

 the "real" image so to exhibit in it details which cannot 

 be seen when this real image is viewed through the ocular 

 of the first microscope. 



This is due not only to the special powers of the sec- 

 ond microscope, but to an advantage which I have taken 

 of a unique fact in photography, namely, that when two 

 lines, markings or colors in an image are too close togeth- 

 er, the sensitive plate will not record them as two but as 

 one. Thus, when I ruled two lines upon a metal plate 

 too closely together, the image of these lines thrown by 

 a camera upon a sensitive plate would irradiate in the 

 film and the picture would show only one line. The line 

 of light falling on the photo-salt in the film spreads by 

 molecular irradiation over more area than the actual 

 width of the line of light, and there is also diffused 

 reflection of this line of light by the semi-transparent 

 substance of the film. To these two causes is due the fact 

 that when the details of two structures are too close 

 together in an image of an object, these structures will 

 photograph as one, and thus the detail will be lost. The 

 line of demarcation between them will, in the film of the 

 sensitive plate, be obliterated by the irradiated and dif- 



