140 SUMMARY OF CURRENT RESEARCHES RELATING TO 



good diatom resolver with its collar in another position ; but when all 

 its zones are tried at once, by the direct illumination, it utterly breaks 

 down. 



With regard to A. pellucida, the strongest resolution is obtained 

 with Powell's vertical illuminator. The long strise can only be seen by 

 this method. Spurious longitudinal striee may be easily seen ; but the 

 true lines are very difficult, and may be estimated to be 120,000 to the 

 in. at the lowest. The transverse I have counted repeatedly, and find 

 them, in Van Heurck's specimens, very constant at 95,000 per inch. 

 The best picture of the trans-strise is obtained with oil-immersion 

 1/12, N.A. 1-4:3, or oil-immersion 1/25, N.A. 1-38, and Powell's oil- 

 immersion condenser, used dry, with single slot, edge of flame direct, 

 valve being dry on cover. The lowest angled glass with which I 

 have seen the transverse striae, is a water-immersion 1/16, N.A. 

 1-08, and the lowest power 1/4, N.A. 1*17." 



In reply to a letter from " Monachus " * inviting Mr. Nelson to 

 state how he came to recognize that oblique illumination must be 

 entirely abolished in favour of central, and that by so doing we shall 

 see the true structure of the object, Mr. Nelson wrote : f — " I began to 

 realize the uselessness of oblique light for the determination of true 

 structure during a lengthened examination of a Nobert's 19-band plate. 

 I was much struck by the appearance of a single line of the first 

 band, when viewed by an oil-immersion N.A. 1*25, illuminated by a 

 large angled cone of direct light. The groove which the diamond 

 had ploughed in the glass was most distinctly seen, and along the 

 sides of the groove there were places where the chips of glass had 

 flown off. With oblique light all this was lost ; the line appeared 

 as if it had been painted on the surface of the glass. This showed 

 me that if definition was wanted direct light must be used. 



I do not intend for one moment to affirm that a higher band of 

 Nobert can be resolved by direct than by oblique light ; but this I 

 do say, that the ultimate structure of a diatom can only be demon- 

 strated by direct light. 



No microscopist in the present day would uphold the theory that 

 the ultimate resolution of the P. angulatum was six sets of lines or 

 grooves, inclined at an angle of 60° to one another. But a similar view 

 of it was held in Quekett's time, for in the frontispiece of his book there 

 is a beautiful engraving of it, exhibiting diamond-shaped marks all 

 over it; a false conclusion, the result of oblique light. Neither 

 will any one insist that the ultimate resolution of the N. Bhomhoides 

 is represented by two sets of lines, at right angles to one another, a 

 picture produced by the employment of two beams of oblique light. 

 In the days of Griffith and Henfrey they got beyond that, and dotted 

 the Bhomhoides. 



It is quite natural to expect that with the increase of aperture 

 and the improvement in objectives there should be simultaneously 

 a development in the resolution of the diatoms. One misses, too, 

 with oblique light, all that beautiful tracery inside the hexagonal 



* Engl. Mech., xxxviii. (1883) p. 341. 

 t Ibid., p. 386 (3 figs.). 



