112 A. A. C. ELIOT MERLIN ON NITZSCHIA SINGALENSIS. 



Thus the conscientious microscopist can no longer feel that he is 

 accomplishing anything requiring great manipulative skill when 

 he obtains distinct rendering of the transverse striae of an average 

 A. pellucida, although of course to show the dotted structure is a 

 far more difficult task. 



With the Nitzschia singalensis the case is different, and any 

 microscopist who can clearly reveal the transverse striation may 

 justly feel satisfied with his optical appliances. I endeavoured 

 to count visually the striae by " stepping " with the cobweb 

 micrometer in the usual way, and noted 116,000 to the inch as 

 the result, but owing to their faintness the eye became rapidly 

 fatigued, and I could not feel any great confidence in the accuracy 

 of the count. Another and more certain method is available, 

 and the magnification at which the photograph (PL 8) was 

 taken was verified as exactly 3,600 diameters by the projection 

 of Grayson's rulings on to the camera screen with the precise 

 optical combination and camera extension used to obtain the 

 photograph. Now the striae on the photograph run at 32 to the 

 inch, which multiplied by 3,600 makes them 115,200 to the inch. 

 This may be relied upon as correct for the specimen photo- 

 graphed. 



In practice I have found that the transverse striae of N. 

 singalensis can be revealed with a good semi-apochromatic 

 objective of 1*33 N.A. by means of axial light and a Gifford screen ; 

 and I should here like to record my conviction, as the result of 

 long experience, that a really well corrected and fine objective 

 will resolve fully as much with axial as with oblique light, although 

 it may need less skill and less keen sight when employing the 

 latter method. I also venture to suggest that, as a test for first- 

 rate objectives of the largest aperture, Nitzschia singalensis 

 must henceforth rank far above A. pellucida, and any failure 

 to resolve the transverse striae of the former with lenses of 

 1*30 N.A. and upwards will indicate a faulty objective or faulty 

 manipulation. Nevertheless, the test is severe, and when accom- 

 plished with axial light requires good eyesight in addition to good 

 optical means. 



