1909.] NATURAL SCIENCES OF PHILADELPHIA. 1S5 



illuminated image. Intermediate forms may appear green or even of 

 a yellowish cast, but never red. 



The agreement in color of the images seen through the left tube with 

 what would be predicated from the examination of spectra admitted, 

 is further evidence that they are due practically exclusively to dif- 

 fracted rays, but still further experiment is required to demonstrate that 

 they are free from refracted rays. The spectra here seen, and which 

 have previously been considered, consist of an infinite number of over- 

 lapping images of the aperture in the subst age diaphragm , each produced 

 by light of a different wave length ; hence only the extreme ends are 

 practically pure colors, the middle portions of the spectra consisting 

 of a jumble of colors sometimes producing the effect of white light; but 

 if a narrow slit diaphragm be inserted in the condenser we can secure 

 a spectrum that corresponds in sharpness to that from an ordinary 

 spectroscope. This may be enlarged by the use of a low power 

 objective in the draw tube, focussed on the back focal plane of the first 

 objective as when using an Abbe apertometer. The microscope may 

 now be used as a spectroscope, and by allowing the light supplied to it 

 to pass through some coloring matter which has well-marked absorp- 

 tion bands, these will be visible in the spectra at back of objective. 

 A solution of Eosin, which will be needed for a later experiment, will 

 likewise answer here. It should be of such strength that when placed 

 in a glass trough and examined with a spectroscope of low dispersion, 

 it will show a black, well-defined band in the green adjoining the 

 yellow, but allow the remainder of the spectrum to pass freely. On 

 placing this trough in the path of the light used to illuminate the 

 microscope, which should be an intense one, it will be noted that the 

 absorption band in the spectrum derived from the diatom structure is 

 perfectly black, furnishing a demonstration that it is practically free from 

 refracted rays; for if it contained scattered refracted rays, as has been 

 claimed, the absorption band would not appear black, but of same color 

 as the light illuminating the object, which in this case is visually a 

 bright red. 



If we now examine the diffraction image produced under these con- 

 ditions, taking the precaution to so adjust the slit diaphragm that the 

 red end of diffraction spectrum is cut off by the margin of the objec- 

 tive's aperture, will find that it is just as blue with this visually red 

 light as it was with white light, owing to the fact that Eosin transmits 

 the blue rays freely; but the midrib of the diatom, and particularly 

 any granular incrustation, such as may usually be found at places 

 between the valve and cover-glass, will be tinted red, indicating that 



