MANAGEMENT OF THE MICROSCOPE. 157 



bear a relation to the structural arrangement on which their production 

 depends. If the Objective be perfectly corrected, and all the diffraction- 

 spectra lie within its field, they will be re-united by the Eye-piece to form 

 a secondary or * diffraction ' image, lying in the same focal plane with the 

 dioptric image, and coinciding with it, while filling up its outlines by 

 supplying intermediate details. But where the markings (of whatever 

 nature) are so closely approximated as to produce a wide dispersion of 

 the interference-spectra, only apart of them may fall within the range of 

 the Objective; and the re-combination of these may produce a diffraction- 

 image differing more or less completely (perhaps even totally) from the 

 real structure; whilst, if they should lie entirely outside the field of the 

 Objective, no secondary or diffraction image will be produced. Thus, 

 whilst the dioptric image represents the actual object, a diffraction- 

 image formed by the reunion of some of the interference-spectra is only 

 an optical expression of the result of their partial re-combination, which 

 may represent something entirely different from the real structure; the 

 same arrangement of lines (for example) being presented to the eye by dif- 

 ferently-liued surfaces, and different arrangements by similarly -lined sur- 

 faces, according to the numbers and positions of the re-united spectra. 1 

 This doctrine, originally based on elaborate theoretical investigations in 

 connection with the ' tfndulatory Theory of Light/ has been so fully 

 borne out by experimental inquiries instituted to test it, and is in such 

 complete harmony with the most certain experiences of Microscopists, that 

 its truth scarcely admits of doubt. Although any attempt to explain its 

 theory in a Treatise like the present must necessarily be altogether futile, 

 yet a selection from the experiments by which Prof. Abbe has verified it, 

 will not only assist in the comprehension of the doctrine, but will enable 

 every Microscopist to satisfy himself of their correctness. 



A 'grating' should be provided, ruled alternately with long and with short 

 lines, as in Fig. 117, A; the lines being traced with a diamond point on a film of 

 silver of extreme tenuity deposited on a thin glass-cover; and the ruled surface 

 being cemented to an ordinary glass slide with Canada balsam. 2 The ' adapter ' 

 ordinarily used for rotating the analyzing prism of the Polariscope between the 

 Objective and the Microscope- body, should be fitted with a small tube for the 

 introduction of diaphragms with varied slits, so that these may be rotated imme- 

 diately behind the back combination of the Objective. The 'grating' being 

 placed on the Stage of the Microscope, illuminated from the mirror, and focussed 

 under a 1-inch Objective, so as to show the ordinary microscopic image of its 

 ruled surface as at A, the eye -piece is removed, and the observer, looking into the 

 body of the instrument, and changing the place of his eyes, sees two rows of 

 spectra, each having a central circle, with ovals on either side of it (C). The 

 central circle is bright and colorless; while each of the ovals shows the colors of 

 the solar spectrum, with the blue always towards the centre. These ovals are 

 'diffraction-spectra;' of which the four closely approximated pairs in the upper 

 row are formed by the wider lines of the single ruling, and the two pairs in the 

 lower row (which are at double the distance of the preceding) by the closer lines 

 of the double ruling. 



The following experiments show (1) that the dioptric image, when viewed by 

 the eye-piece separately from all diffraction-spectra, gives no Microscopic repre- 



1 The reader may, perhaps, be aided in comprehending Prof. Abbe's doctrine 

 by the following analogy: When a solar spectrum is projected by a prism on a 

 white surface, its entire re-combination by a convex lens will reproduce a beam 

 of white light. But, if only certain parts of the spectrum be thus recombined, 

 the beam will have a color dependent upon the selection. 



2 In the grating used by Mr. Stephenson (" Monthly Microsc, Journ.," Vol. 

 xvii., p. 83), the lines in the upper half were about 1,790 to the inch, and in the 

 lower about 3,580. 



