^6 ON DIFFRACTION SPECTRA IN THE MICROSCOPE. 



situate pencils are seen in the aperture image almost in its diameter 

 (indicated in pi. III., fig. 19, where the dotted circle shews the 

 shifted image with oblique illumination, and c, g are the two not 

 next situate pencils lying nearly in diametral position) . If a card 

 diaphragm be notched, as in fig. 19, that is to say, the passage of the 

 pencils c, g, permitted, by cutting out the pieces marked by the small 

 arcs Jc, I, and this card affixed behind the back lens of the objective 

 (to the adapter) the phenomena of this second series of lines may 

 be produced. A good object is necessary j that is, a flat Pleuro- 

 sigma valve with strong markings ^mounted dry) and a good 

 immersion, ith or -j^th, accurately corrected for thickness of 

 cover.* (2ndly) Another series, exactly similar to the known 

 hexagonal pattern, must become visible when the three pencils, 

 J, d, or €, e, g, are brought into action together, all other pencils 

 being shut off. These lines have also interspaces smaller in the 

 ratio of ^3 • i than the original series, and the sides of the 

 hexagonal fields would show at right angles to the midrib of the 

 valve. (3rd J lastly) Three additional systems of striae, taking the 

 same direction as the original striae, but of exactly double fineness, 

 would be produced by using the paired pencils, I, e^ e, f, d, g. 

 These series (2 and 3) can be only successfully obtained by using 

 direct light of the sun. 



In conclusion, the writer would remark that Professor Abbe's 

 arrangement of a diaphragm holder rotating in an adapting ring, 

 besides being simple and inexpensive, has this special excellence, 

 that it enables all the changes necessary for producing or with- 

 drawing diffraction images to be effected instantaneously without 

 any alteration of focus or manipulation of the instrument during 

 the experiments. The conclusions at which the observer's mind 

 arrives are thus unembarassed by any suspicion of changed circum- 

 stances other than those intended and calculated. 



* By the same means we may obtain, among other combinations, line 

 systems parallel to the axis (midrib) of the valve through the combination 



C,g,OTb,f. 



