582 Dr. G. Johnstone Stoney on 



of the medium in which the microscopical object is mounted ; 

 so that the reader is requested to picture all three — stratum 

 S, the cover-glass, and the mounting medium— as trans- 

 parent and of unlimited extent horizontally. 



32. In order to be able to employ the Method of Reversal 

 in our investigation, we here make the legitimate hypothesis 

 that the air within stratum S is replaced by a medium which 

 is perfectly transparent, and which we may call transparent 

 air. We may further suppose the apparatus that is over this 

 air space to be removed — the microscope tube and all it 

 carries — and that transparent air fills the entire space over 

 stratum S, which we shall call space K. Under these circum- 

 stances the light from the microscopical objects, i. e. from the 

 microscopical object and its surroundings, will travel un- 

 interruptedly forward, crossing stratum S and advancing 

 into K, the added air space above it. Let us suppose that 

 the emission of light from the microscopical objects has been 

 allowed to go on for some definite time, say for one minute. 

 At the end of the minute let Reversal of the light take place, 

 the effect of which will be that the light which had been 

 travelling upwards through space K, will now T retrace its 

 steps, will recross stratum S, will re-enter the cover-glass, 

 will pass from it into the medium in which the microscopical 

 object has been mounted, and will finally converge upon the 

 microscopical objects from which it had emanated. This is 

 what would happen if Reversal alone took place; but we shall 

 find it useful to suppose that at the instant of Reversal, the 

 microscopical object is removed and mounting material put in 

 its place. Under these circumstances the light upon its return 

 journey after Reversal, would recross stratum S and the 

 cover-glass, and advancing into the mounting medium would 

 in it converge upon and form an image of the microscopical 

 object, which image is entitled to the name of its standard 

 image, inasmuch as, although inconveniently small, it is the 

 most perfect image of that object that the light in question is 

 competent to form, i. e. the most perfect as regards fulness of 

 detail and accuracy of delineation. We shall give to it the 

 name of Image B'. 



33. A slight modification of the procedure w r e have followed 

 will furnish another standard image, which sball include pre- 

 cisely the same detail as Image B', and will be formed under 

 conditions that are more convenient for the use we intend to 

 make of it. To form this image, proceed as before until the 

 First Reversal has been effected in space K. This, as before, 

 causes the light to retrace backwards the course which before 

 the reversal it had pursued upw r ards. At the instant of 



