E. M. NELSON ON ACTINOCYCLUS RALFSII. 379' 



colour, which like all diffraction colours turns white on resolution, 

 or more strictly speaking shortly before resolution. 



The tint of the diffraction colour of a diatom depends upon 

 (a) the aperture of the objective used, and (/?) the obliquity 

 of the illumination. By this means we may therefore roughly 

 determine the fineness of any diatomic structure by matching 

 the tint with one whose fineness of structure has been measured, 

 or with a test plate of ruled bands. 



Of course it is necessary that the comparison be made with 

 the same objective and under the same conditions of illumination. 

 A suitable illumination for this purpose is daylight, and an 

 achromatic condenser with a central opaque stop, just large 

 enough to give a dark ground. 



The question then is : what is the cause of the colour in 

 Actinocyclus Ralfsii ? Obviously it cannot be a diffraction colour 

 arising from the ordinary primary structure forming the " rays," 

 which give the diatom its name, because as we have seen 

 above, when this structure is resolved the colour is still visible, 

 and no colour arising from diffraction is visible when the 

 diffractor itself is resolved. It cannot be due to pigment, for 

 if it were it would remain visible when the aperture was 

 increased beyond *45 N.A. It cannot be caused by thin plates, 

 because it would require reflected and not transmitted light 

 to render it visible. Polarisation and refraction seem quite out 

 of the question ; and as there is no other theory at hand the 

 answer must for the present be left undetermined. 



It was pointed out in 1897 {Journ. Q. M. C, Vol. 6, ser. 2, 

 p. 431) that with an apochromatic J of 1*4 N.A., used in 

 connection with a wide-angled oil-immersion condenser giving a 

 large aplanatic cone, a very delicate perforated veil could be seen 

 covering the w^hole valve of an Actinocyclus Ralfsii. This very 

 delicate structure has obviously nothing to do with the colour 

 in question, because it would require a far greater aperture 

 than "45 N.A. to develop upon a dark ground any colour arising 

 from the diffraction of so fine a grating ; and this question is 

 quite independent of that concerning the different kind of 

 illumination required to develop the colour, a point of which 

 we have as yet found no explanation. If a Hyalodiscus suhtilis 

 whose structure is about 70,000 per inch, or twice as fine as 

 that of Hyalodiscus stelliger, be examined on a dark ground 



