1909.] ' NATURAL SCIENCES OF PHILADELPHIA. 183 



with low powers, as no wide-angled objectives have been made that 

 are sufficiently well corrected to stand a full cone without breaking 

 down, unless on objects such as deeply stained bacilli, where the image 

 is formed principally by absorption and is practically a silhouette. 

 There seems to be no reason why such images should not be regarded 

 as dioptric, although there must be some diffracted rays from the 

 margins which are undoubtedly utilized in the image. 



When the back of the objective is entirely filled with light, the study 

 of diffraction phenomena becomes difficult, as even with particularly 

 suitable objects the diffraction beams are eclipsed by the brighter 

 dioptric beam. As we open the iris of the condenser, however, it can 

 be seen that the diffracted beams expand to the same extent as the 

 dioptric beam, finally overlapping it and each other, until when the 

 aperture of the objective is entirely filled with dioptric light it must 

 unquestionably be similarly filled with diffracted rays. Unfortu- 

 nately, there seems to be no way in which we can completely separate 

 the beams derived from the two different sources, and the best expe- 

 dient I could devise to obtain some idea of the conditions present con- 

 sists in the use of a semicircular diaphragm in the condenser, so oriented 

 that the left side of the back of objective is completely filled with light 

 up to its margin, while the right side receives no direct rays whatever. 

 Thus we have full cone conditions on one side, while the other will 

 receive only refracted and diffracted rays, or if certain suitable objects 

 are employed nothing but diffracted rays, whose behavior we can 

 study separately. 



For this purpose a binocular microscope should be employed with a 

 specially short mounted objective, say a sixth of about .80 N. A., the 

 back lens of which will come close to the Wenham prism. All the 

 direct light from the fully illuminated left half of the objective will now 

 pass up the right-hand tube of the microscope, while the diffracted 

 beams from the right-hand half of objective will be reflected up the 

 left tube. Assuming that P. angulatum is again the object, as it fur- 

 nishes spectra singularly free from all indications of refracted light, and 

 oriented longitudinally across the field in a right-and-left direction, on 

 examining the back of objective, the previously dark space on the 

 right will be found practically filled by three of the characteristic 

 spectra of the object, and the other three will be present, although 

 invisible, in the illuminated half, where they will occupy positions near 

 the margin, as they can only be derived from central or nearly central 

 dioptric rays. The diffracted beams in right half of objective, being- 

 derived from the dioptric beam which completely fills the left half 



